Price Transmission and Market Integration of Fish in Oyo State
MARKET INTEGRATION AND PRICE TRANSMISSION ANALYSIS …
Transcript of MARKET INTEGRATION AND PRICE TRANSMISSION ANALYSIS …
MARKET INTEGRATION AND PRICE TRANSMISSION ANALYSIS OF PROCESSED
FISH MARKETS IN GHANA
BY
VICTOR NTOW DJARBENG
(10373751)
THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON, IN
PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE AWARD OF MASTER
OF PHILOSOPHY DEGREE IN AGRICULTURAL ECONOMICS
DEPARTMENT OF AGRICULTURAL ECONOMICS AND AGRICUSINESS, SCHOOL
OF AGRICULTURAL, COLLEGE OF BASIC AND APPLIED SCIENCE UG, LEGON
JULY, 2018
University of Ghana http://ugspace.ug.edu.gh
i
DECLARATION
I, Victor Ntow Djarbeng, author of this thesis titled “Market Integration and Price
Transmission Analysis of Fish Markets in Ghana” do here by declare that with the
exception of the references duly acknowledged; this work was undertaken by me from
August 2017 to July 2018 in the Department of Agricultural Economics and
Agribusiness, University of Ghana, Legon. I do hereby declare that, this work has not
been submitted in part or whole for a degree in this University or anywhere.
--------------------------------------------
Victor Ntow Djarbeng
Date: -----------------------------
This thesis has been presented for examination with our approval as supervisors
------------------------------ --------------------------------
Rev. Dr. Edward Ebo Onumah Prof. Al-Hassan Wayo Seini
(Major supervisor) (Co-supervisor)
Date: ----------------------- Date: -----------------------
University of Ghana http://ugspace.ug.edu.gh
ii
DEDICATION
I dedicate this thesis to my entire family for the care, love and support they have shown
me throughout my education. May the Grace and Favour of our Lord Jesus Christ always
be with you.
University of Ghana http://ugspace.ug.edu.gh
iii
ACKNOWLEDGEMENT
My utmost gratitude goes to God Almighty for his grace, mercies and favour which has
seen me through to the completion of this study.
I would like to extend my special gratitude to my Major supervisor, Rev. Dr. Edward Ebo
Onumah for his invaluable advice constructive comments, criticisms, directions,
contributions, encouragement and also for his help in securing the funding for the project
work. His role as my supervisor in the research and writing of this thesis was outstanding.
I am also grateful to Prof. Al-Hassan Wayo Seini for his contributions and constructive
comments to my thesis. I acknowledge with gratitude the roles played by all the lecturers
of the Department of Agricultural Economics and Agribusiness, University of Ghana,
Legon and for their immense contributions to this work.
I am very much grateful to Professor Henry De-Graft of the School of Agriculture,
University of Cape Coast for his invaluable advice, constructive comments and immense
help in shaping up my research objectives and methodologies. I am especially thankful to
the Fish4Food project for providing the funding required to successfully complete this
thesis. I would like to thank the Ghana Statistical Service for providing the price data sets
used for the study.
My sincere gratitude goes to my siblings Judah and Dromo Djarbeng for their love,
prayers and support.
Last but not least, my sincere thanks go to all who played some role in one way or the
other, but whose names have not been mentioned. May the good Lord bless you all.
Victor Ntow Djarbeng
University of Ghana http://ugspace.ug.edu.gh
iv
ABSTRACT
This study presents an analysis of market integration and price transmission of 10
processed fish markets in Ghana over the period 2012-2017. Monthly processed fish
prices for Koobi, Kpala and Smoked herrings from the 10 markets namely Makola, Ada,
Agbozume, Cape Coast, Half Assini, Kpandu, Mankessim, Techiman, Tema and Wa
markets were used for the analysis. Data was analysed using the Johansen Co-integration
procedure, Granger Causality test, Asymmetric Vector Error-Correction model and
Autoregressive Distributive models. The results revealed a lack of cointegration among
most of the markets for Kpala, smoked herrings and koobi indicating that their markets
were not linked together. The granger causality test showed unidirectional, bidirectional
and independent causality in the Koobi and Kpala markets. The smoked herring markets
exhibited only bidirectional and unidirectional causality. The speed of adjustment from
the asymmetric vector error correction model for cointegrated markets where higher for
smoked herring than for Koobi. The study further revealed that positive shocks were
corrected faster than negative shocks. There was asymmetry present only in the Makola-
Mankessim markets for smoked herring. The results of the autoregression distributive lag
model show that most of the current and previous prices of the markets did not influence
the current prices in the Makola market. Thus, in most cases the price of processed fish
was dependent on the prevailing market conditions. In conclusion, the results show the
presence of poor price transmission and poor integration between processed fish markets.
It is recommended that infrastructure could be improved to enhance market integration
and efficiency. Also information on the prices of processed fish should be made available
to processers to help them know which markets offer lucrative prices for their product.
This could improve spatial arbitrage and hence market integration.
University of Ghana http://ugspace.ug.edu.gh
v
TABLE OF CONTENTS DECLARATION .............................................................................................................................. i
DEDICATION ................................................................................................................................. ii
ACKNOWLEDGEMENT .............................................................................................................. iii
ABSTRACT .................................................................................................................................... iv
LIST OF TABLES .......................................................................................................................... ix
LIST OF FIGURES ......................................................................................................................... x
LIST OF ACRONYMS .................................................................................................................. xi
CHAPTER ONE .............................................................................................................................. 1
INTRODUCTION ........................................................................................................................... 1
1.1 Background to the Study ............................................................................................................ 1
1.2 Problem Statement ..................................................................................................................... 6
1.3 Research Questions .................................................................................................................... 8
1.4 Research Objectives of the Study .............................................................................................. 9
1.5 Justification of the Study ........................................................................................................... 9
1.6 Organization of the Study ........................................................................................................ 10
CHAPTER TWO ........................................................................................................................... 11
LITERATURE REVIEW .............................................................................................................. 11
2.0 Introduction .............................................................................................................................. 11
2.1 Economic Importance of Fish in Ghana .................................................................................. 11
2.1.1 Contribution of Fish to Food Security .................................................................................. 11
University of Ghana http://ugspace.ug.edu.gh
vi
2.1.2 Contribution of Fish to Poverty Reduction ........................................................................... 12
2.1.3 Contribution of Fish to Employment .................................................................................... 12
2.3 Processing of Fish .................................................................................................................... 15
2.4 The Concept of Markets........................................................................................................... 16
2.5 The Concept of Market Integration, Price Transmission and Spatial Arbitrage...................... 18
2.5.1 Price Transmission ................................................................................................................ 20
2.5.2 The Concept of Spatial Arbitrage ......................................................................................... 21
2.5.3 The law of One Price ............................................................................................................ 22
2.6 Conditions for Efficient Market Arbitrage .............................................................................. 24
2.7 Techniques for Measuring Spatial Market Integration ............................................................ 24
2.8 Asymmetry in price transmission ............................................................................................ 30
2.8.1 Causes of Asymmetry in Price Transmission ....................................................................... 31
2.9 Empirical Studies linked to Market Integration and Price Transmission in Ghana and
Elsewhere ....................................................................................................................................... 33
2.8 Summary .................................................................................................................................. 40
CHAPTER THREE ....................................................................................................................... 41
METHODOLOGY ........................................................................................................................ 41
3.0 Introduction .............................................................................................................................. 41
3.1 Theoretical Framework ............................................................................................................ 41
3.2 Conceptual Framework ............................................................................................................ 42
University of Ghana http://ugspace.ug.edu.gh
vii
3.3 Empirical Framework .............................................................................................................. 43
3.4 Method of Analysis .................................................................................................................. 46
3.4.1Analyzing the extent of market integration ........................................................................... 46
3.4.1.1 Johansen Cointegration ...................................................................................................... 47
3.4.1.2 Test for Granger Causality ................................................................................................. 49
3.5 Extent of Price Transmission ................................................................................................... 51
3.5.1 Extent of Price Transmission in Processed fish Market (ARDL) ......................................... 51
3.5.2 Extent of Price Transmission in Processed fish Market (VECM) ........................................ 52
3.6 Estimation of Extent of Asymmetry in processed fish Market ................................................ 53
3.7 Study Area and Data Source .................................................................................................... 55
CHARPTER FOUR ....................................................................................................................... 57
RESULTS AND DISCUSSIONS .................................................................................................. 57
4.0 Introduction .............................................................................................................................. 57
4.1 Descriptive Analysis of Processed Fish Markets ..................................................................... 57
4.2 Unit Root Test Results ............................................................................................................. 63
4.2.1 Unit Roots Test Results for Processed Fish .......................................................................... 63
4.3 Cointegration Test Results ....................................................................................................... 67
4.3.1 Extent of Market Integration between Processed Fish Market ............................................. 67
4.4 Direction of causality in Processed Fish Market ..................................................................... 70
4.4.1 Result of Granger-causality Test for Processed Fish Markets .............................................. 70
University of Ghana http://ugspace.ug.edu.gh
viii
4.5 Price Transmission between Processed Fish Markets in Ghana. ............................................. 72
4.5.1 Results of the Asymmetric Vector Error Correction Model ................................................. 72
4.5.2 Evidence of Price Transmission (ARDL Model) .................................................................. 75
CHAPTER FIVE: .......................................................................................................................... 78
SUMMARY, CONCLUSSIONS AND RECOMMENDATIONS ............................................... 78
5.0 Introduction .............................................................................................................................. 78
5.1 Summary and Major Findings ................................................................................................. 78
5.2 Conclusions .............................................................................................................................. 80
5.3 Policy Recommendations......................................................................................................... 81
APPENDICES ............................................................................................................................... 99
University of Ghana http://ugspace.ug.edu.gh
ix
LIST OF TABLES
TABLE PAGES
Table 4.1 Descriptive statistics of monthly processed fish prices (2012-2017) ................ 60
Table 4.2: Results of ADF tests on the monthly processed fish price series ..................... 66
Table 4.3 Johansen cointegration test Statistics for processed fish markets ...................... 69
Table 4.4: Result of granger-causality test for processed fish markets ............................. 71
Table 4.5 Results of the AECM Model ............................................................................. 75
Table 4.6 Results of the ARDL model............................................................................... 77
University of Ghana http://ugspace.ug.edu.gh
x
LIST OF FIGURES
FIGURE PAGES
Figure 3.1 Empirical framework for assessing price transmission....................................44
Figure 3.2: Map of Ghana showing the locations of the market under study....................56
Figure 4.1: Trend plot of monthly prices of Koobi............................................................62
Figure 4.2: Trend plot of monthly prices of Kpala............................................................ 62
Figure 4.3: Trend plot of monthly prices of smoked herrings............................................63
University of Ghana http://ugspace.ug.edu.gh
xi
LIST OF ACRONYMS
ADF Augmented Dickey Fuller
APT Asymmetric Price Transmission
ARDL Auto Regressive Distributed Lag
AVECM Asymmetry Vector Error Correction Model
CPI Consumer Price Index
ECM Error Correction Model
ECT Error Correction Term
ESTJ Enke-Samuelsson-Takayama-Judge
FAO Food and Agriculture Organization
FASDEP Food and Agricultural Sector Development Plan
FIFO First-In-First-Out
GSS Ghana Statistical Service
KPSS Kwiatkowski, Phillips Schmidt and Shin
LOP Law of One Price
METASIP Medium Term Agriculture Sector Investment Plan
PBM Parity Bound Model
TAR Threshold Autoregressive Model
VAR Vector Autoregressive Model
VECM Vector Error Correction Model
University of Ghana http://ugspace.ug.edu.gh
1
CHAPTER ONE
INTRODUCTION
1.1 Background to the Study
Markets have an important role in facilitating the interactions between the forces of
demand and supply such as through the adjustment of commodity prices across time and
space and also in dealing with the risks arising as a result of shocks to the demand and
supply factors existing in a market. Markets that are well-integrated tend to facilitate
adjustments in net export flows across space, thereby reducing variability in prices
encountered by both consumers and producers (Barrett, 2005).
In developing countries, the degree to which markets are integrated and the price
transmission of shocks to food markets is a key determinant of stability in prices and
overall food security. In markets that are not well integrated signals from price shocks
may not be passed on from food deficit areas to food surplus areas resulting in prices
becoming increasingly volatile which might affect the purchasing power of consumers
and hence their economic accessibility to food. The existence of poor market integration
among spatially separated markets may also end with a decrease in the information on
prices accessible to economic agents; this may limit the allocative efficiency and long run
growth which in turn affects the availability and supply of food. According to Onyuma et
al. (2006), most agricultural markets in Africa countries are poorly integrated and not
efficient.
Spatial market integration can be used to indicate efficiency between markets that are
situated in regions or location and this is important in closing the disparity gap, improving
food security by ensuring food is readily available from food-surplus to food-deficit
University of Ghana http://ugspace.ug.edu.gh
2
regions, improving upon the standards of living of rural poor households, living standards
of poverty stricken rural households, improving adoption of technologies and
implementing effective macro-level policies; that ensure that welfare gains from
implemented policies are realized (Ankamah-Yeboah, 2012). Perfect integration of
markets and complete transmission of price shocks, with immediate adjustment to
variation in prices caused by shocks from within or external to the system is essentially a
marketing system that can be described as efficient. A system such as this would allow
the market agents in the marketing chain to receive the greatest benefits. It would also aid
in removal of unprofitable arbitrage and segmentation of markets that are separated by
location and would make certain that allocation of resources across space and time is
efficient (Nkang et al., 2007). A farm marketing system that is efficient helps to increase
the revenue levels of farmers and also promote the economic development of a nation
(Aburajab, 1999).
Integration of markets measure the degree to which spatially separated markets for a
homogenous commodity in the long run share common price or trade information. Spatial
integration of markets improves successful trade between areas experiencing shortages in
food supply and areas with an abundance of food. This results in specialization and
economic growth. Integration of markets contributes greatly to food security and growth
of the economy; it also enhances the social welfare of economic agents and speeds up the
rate of effective transmission of changes in between markets with the support of market
reforms (Goletti & Babu, 1994). However, poor integration of markets on the other hand
reveals the presence of inaccurate information on prices, existence of government
University of Ghana http://ugspace.ug.edu.gh
3
policies, problems of infrastructure or institutional problems that affect the market
decision of producers and the efficient flow of goods from one market to the other.
Price transmission of domestic regional markets is central in understanding the degree to
which economic agents are integrated into the markets process. The nonexistence of
integration among markets or the complete pass-through of changes in price from one
market to the other has important repercussions for the welfare of the economy.
Incomplete transmission of prices between markets that may arise due either to the cost of
transactions or to trade and other policies, results in a decrease in information on prices
accessible to economic agents resulting in decisions that lead to inefficient outcomes.
Price transmission studies also help to provide understanding on how changes in prices
are transmitted between markets and as such indicating the degree to which markets are
integrated, as well as the degree to which these markets function efficiently.
To a large extent the performance of the agriculture sector depends on production or
supply efficiency and also on marketing efficiency, chiefly for agricultural markets and
price signal. Agricultural prices and hence agricultural markets significantly affect the
speed and path of developments in agriculture, and given the significance of the
agricultural sector to the Ghanaian economy, an increase in the level of efficiency of
Ghanaian agricultural markets will help to improve upon the growth rate of the economy.
Over the years FASDEP has concentrated on modernizing agricultural markets, so as to
create connections in the value chain and highlights the use of resources with
sustainability in mind and commercializing agricultural activities promote expansion of
markets (FASDEP II, 2002; FASDEP II, 2008). Agriculture in Ghana is very important
and as such a high level of efficiency in agricultural markets in the country will be
University of Ghana http://ugspace.ug.edu.gh
4
relevant for the development of the economy. The sustained efforts to develop and
modernize markets for agriculture over the past years following the arrival of Ghana as a
middle income country raises issues on the present state of performance and response of
markets located at different location to each other.
Isolated or segmented markets may receive incorrect information on price changes that
might alter the marketing decision of producers and cause the inefficient movement of
commodities (Alderman & Shively, 1991). Markets that are integrated allow deficits or
surpluses in a market to be passed on to another market through arbitrage; an
improvement in the integration of markets situated at different locations will bring about
a balance among food–deficit and food-surplus regions. In addition, it will ensure that
prices of goods in markets situated at different locations follow each other and that price
signals and information are passed on smoothly. As such governments may be interested
in knowing how well their markets are functioning and the price movement relationships
of staple foods in the various ecological zones as this helps to improve food security in
the country by identifying areas in which government can concentrate its efforts.
Although a lot of studies have examined the integration of market integration and
transmission of shocks to price in Ghana, most of them have focused on grains and
cereals especially maize neglecting other important commodities such as fish which are
also important to food security and development of the nation.
Fish plays a major role in supplying of animal protein and micronutrients to meet the
protein requirement of human beings all over the world (Quaas et al., 2016). It is
responsible for providing about 4.5billion consumers with at least 15 % of their average
University of Ghana http://ugspace.ug.edu.gh
5
per capita intake of animal protein (Béné et al., 2015). Fish is also more nourishing than
most staple foods such as cereals, providing in particular essential fatty acids and
micronutrients and as such is very significant in improving the nutritional status of
individuals, in particular those at risk such as children and women (Béné et al., 2015).
The unique nutritional properties of fish make it essential to the health of billions of
consumers in both developed and developing countries (Béné et al., 2015). Fish is one of
the most efficient converters of feed into high quality food and its carbon footprint is
lower compared to other animal production systems (Béné et al., 2015). With the
increasing risk of becoming infected with a variety of diseases associated with meat
consumption, more and more people are depending on fish to meet their protein needs.
The annual per capita consumption of the world for fish has risen from 13 kg-19 kg
between 2000 and 2012, respectively (FAO,2013). Demand for fish is expected to rise
given the rapid increase in human population and the subsequent increase in demand for
animal protein especially in developing countries.
It is estimated that fish contribute about 3% to the gross domestic product (GDP) and 5%
to the agriculture GDP of the nation (Awity, 2005). The Fisheries sector in Ghana directly
employs about 2.2 million people in the nation and contributes 4.4% to the GDP of the
nation (Anon, 2008). This shows an increase in GDP from that reported by Awity (2005).
Currently the sector accounts for 6.1% of the agricultural GDP (MOFA, 2016). In Ghana
the average per capita fish consumption is around 20-25kg, which is higher than the
world and Africa average of 19 kg and 10 kg, respectively average of 13kg (Anon, 2008).
Fish still remains the number one and less expensive source of animal protein and making
up about as much as 60 % of animal protein requirement in the diet of Ghanaian (FAO,
2016). This represents about 22.4 % of household expenditures spent on food. Most
University of Ghana http://ugspace.ug.edu.gh
6
Ghanaians prefer to consume fish in smoked form because of its taste and delicacy. Other
methods for preserving and processing fish include drying, frying, salting and fermenting
just to name a few.
In developing countries such as Ghana more than 80% of fish produced in the country is
consumed as food while the leftover of 20% or less are used for other activities such as
preparing feed for some animals and fish oil. As such it becomes necessary to develop an
efficient network of marketing system to in the nation so that a large amount of the fish
produced in the country can be efficiently managed and supplied to consumers, with the
benefits of the fisherman in mind.
1.2 Problem Statement
Fish remains an important source of food, income and livelihood for millions of people
around world including Ghana. Due to the continuous increase in the population of the
country and the subsequent increase in the demand for fish combined with the decreasing
fish stock may lead to food insecurity. One way to solve the problem of food insecurity is
through the implementation of policies that improve the market efficiency and stability of
prices in fish markets. There is however no empirical information on degree of the
integration of market and transmission of changes in the price of fish to inform the
designing of such policies.
The spatial market integration and price transmission of signals has become a general
way for evaluating the performance of market in some countries. Spatial integration of
markets looks at the degree to which markets located at different locations have
information on price or trade of a similar commodity that are common to both markets in
University of Ghana http://ugspace.ug.edu.gh
7
the long-run. These markets are linked by arbitrage and this ensures that goods move
from a region of high production to a region of low or no production. The importance of
information on the integration of markets is in its use in formulating policies, and, on the
extent to which market development may be promoted. Market integration will also help
to understand the movement of equilibrium paths of demand and supply for processed
fish. Segmented markets are more likely to transmit incorrect price information that might
distort the marketing decisions of producer that may contribute to inefficiencies in the
movement of goods. Such information is important in formulating policy strategies to
prevent food insecurity (Goletti & Babu, 1994).
The movement of prices in the country could provide insights into on how the variations
in the price of a good in one market can influence the price, output, consumption and
social welfare of the same commodity in different market. This will also provide
information on how producers and consumers will react to prices changes in another
market.
Another concern of interest to stakeholders when dealing with the responses of markets to
one another is whether the adjustment process between markets is characterized by
symmetric or asymmetric relationships due to the possible influence of traders on market
conduct. Ben-Kaabia et al. (2002) show that relationships characterized by symmetry can
often be taken to represent a case of competitive markets, while relationship that are
asymmetric in nature relationships on the other hand can be linked with the presence of
some inefficiencies present in the market. The existence of asymmetry in the transmission
of changes in price means that a number of market agents are experiencing a loss in
welfare since the distribution of welfare under symmetric conditions could be different
University of Ghana http://ugspace.ug.edu.gh
8
(Wlazlowski et al., 2009). Thus, under conditions of asymmetry in transmission of price
shocks, the allocation of welfare effects across levels and among agents as a result of
shocks in the marketplace will be distorted relative to the situation of price transmission
under conditions of symmetry. On the other hand, adjustments to changes in prices under
symmetry will result in efficient use of resources, increased revenues, enlarge markets for
goods and service, indicate the extent of competitiveness, create jobs and promote value
(Acquah & Owusu, 2012). According to Peltzman (2000) transmission of prices
characterized by asymmetry is common in most markets, but says that, any economic
analysis that does fails to account for the presence of asymmetry in price transmission is
inadequate.
Most of the fish produced in the country is consumed in its processed form (smoked, fried
and dried). It is in the light of this that the study seeks to carry out an analysis of price
transmission and integration of processed fish markets in the country. This is because of
the usefulness of such information in designing agricultural policies aimed at the
stabilization and risk management of price and also food security.
The study does this by asking the following question.
1.3 Research Questions
1. What is extent of market integration among the selected processed fish markets in
Ghana?
2. What is the extent of price transmission among selected processed fish markets in
Ghana?
University of Ghana http://ugspace.ug.edu.gh
9
3. What is the extent of price asymmetry in selected processed fish markets in
Ghana?
1.4 Research Objectives of the Study
The main objective of this study was to examine the efficiency of processed fish
marketing system in Ghana
The specific objectives are to:
1. Analyse the extent of market integration among the selected processed fish
markets in Ghana.
2. Analyse the extent of price transmission among selected processed fish markets in
Ghana.
3. Analyse the extent of price asymmetry in selected processed fish markets in
Ghana.
1.5 Justification of the Study
The examination of the extent of integration of market and transmission of processed fish
prices in Ghana is important. This is because results of the extent of market integration
will be useful in evaluating if processed fish markets are performing their function
effectively and efficiently.
Information relating to the dynamics of price movement between the producer and the
consumer markets may make policy formulation and implementation more successful as
it provides information on the time taken or adjustment period for price policy to be
transmitted across markets. The results of the study will also provide useful information
University of Ghana http://ugspace.ug.edu.gh
10
in the designing of policies and measures aimed at improving price transmission between
markets and hence market efficiency.
Information on the extent of asymmetry in the transmission of prices in the processed fish
sector will be important in understanding whether consumers and producers are
benefitting from price decrease and price increases respectively.
Also most price transmission studies have focused on cereals crops neglecting other food
sources such as fish which also feature in the daily diet of the average Ghanaian. This
study will add to the growing literature on the integration of market and transmission of
price shocks of other commodities.
1.6 Organization of the Study
This study is divided into five chapters. Chapter one provides an introduction to spatial
integration of markets and price transmission analysis. This chapter also presents the
problem statement, research objectives and justification of the study. Chapter two
presents a review of literature on the economic importance of fish in Ghana, theoretical
concept and empirical issues relevant to the studies on integration of market, models for
estimating transmission of price changes and empirical evidence linked to integration of
market and asymmetry in price transmission. Chapter three provides an idea of the
theoretical framework, conceptual and empirical framework in price transmission, and the
methods of analysis used in the study of price transmission that employs times series data.
Chapter four present the result of the study while chapter five presents the summary and
conclusions and policy recommendations of the study.
University of Ghana http://ugspace.ug.edu.gh
11
CHAPTER TWO
LITERATURE REVIEW
2.0 Introduction
This chapter provides a brief review of the various concepts and literature of importance
to the study is presented in this chapter such as the theoretical concepts and empirical
evidence related to agricultural price transmission studies.
2.1 Economic Importance of Fish in Ghana
The fishing sector in Ghana contributes considerably to the socio-economic development
of Ghana in terms of creation of jobs, poverty reduction in both rural and urban
communities and food security in the form of protein intake.
2.1.1 Contribution of Fish to Food Security
In terms of its contribution to food security, fish has been identified as the most important
source to meet the animal protein needs of the Ghanaian population (Aggrey-Fynn, 2001).
According to Kawarazuka & Béné (2011), fish is rich in Omega-3 fatty acids, minerals
and micro-nutrients that can help to bring down blood pressure and aid in minimizing the
possibility of having a heart attack or stroke. It is consumed by most Ghanaians in both
urban and rural areas and supplies consumers with about 60 per cent of their animal
protein requirement. The annual per capita consumption of fish was 24kg in 2014 and this
has increased to 28kg, well above the global average of 18.9kg and 10.5kg for the African
continent (Robadue et al., 2018). Various species of fish are available in their fresh or
processed form to suit the taste, preferences and buying power of the consumer. In
Ghana, fish is commonly smoked or dried to improve its shelf life; it also becomes easier
to transport the finished product to areas far away from production zones in coastal areas.
University of Ghana http://ugspace.ug.edu.gh
12
Small pelagic represent the most important locally produced fish supply and are an
affordable and accessible protein source for poor households (Robadue et al., 2018).
2.1.2 Contribution of Fish to Poverty Reduction
The fisheries sector also helps to reduce poverty in Ghana and this is important in the
development and growth of the Ghanaian economy. Many poor and vulnerable people
especially in the rural areas depend in one way or another on the fisheries sector for their
livelihoods. The fishing sector offers a wide range of full-time or part time livelihood
opportunity to many Ghanaians such as fishing, processing, transporting and marketing of
fish (Mensah, 2012). Through these fish-related activities (fisheries and aquaculture but
also processing and trading), fish contribute substantially to the income and therefore to
the indirect food security of more than10 % of the world population, essentially in
developing and emergent countries (Béné et al., 2015). According to Mensah et al.
(2001), fisheries activities around the Volta Lake serve as a fall-back livelihood strategy
for many migrants to the area. A study by Ofori-Danson (2013) showed that the poverty
head count index for fishing communities in Ghana ranged between 60-80 percent for
inland areas and 50-72 percent for coastal areas.
2.1.3 Contribution of Fish to Employment
The fishing industry in Ghana employs about ten percent of the total population in the
country (FAO, 2016). Men in the fishing industry are mostly involved in the harvesting of
fish and also undertake the main fishing activities in the various subsectors whereas
women are mostly involved in post-harvest activities such as fish processing and storage
and also in the marketing of fish. Many are also engaged in the frozen fish distribution
trade as well as fish exports. The artisanal subsector supplies over 70 percent of the fish
University of Ghana http://ugspace.ug.edu.gh
13
produced in the country and has generated jobs for over 60 percent of the female
population involved in the value chain for fishery (Bank of Ghana 2008). Fishers in the
artisanal subsector number about 120 000. The marine subsector employs over 135 000
fishers including makers of the boats used in fishing, office workers for industrial fishing
fleets and input suppliers (FAO, 2016). Those engaged mainly in the processing and
distribution of fish are judged to be about 500 000 in number. The number of people
depending directly on the Volta Lake for their daily survival numbered around 300 000
(FAO, 2016). Out of this number 26.66 percent representing 80 000 are fishermen and the
rest are fish processors, traders (FAO, 2016). Some people are also involved in the
packing, storing, loading, unloading and transporting of both fresh and processed fish and
fish products. Others jobs also include the cannery workers, fishmeal producers and their
staff and export processors.
2.2 Fish Marketing
In recent years, fish marketing in Ghana has become important due to the growing
awareness of the health benefits derived from eating fish. Fish marketing in Ghana is
mostly dominated by women in both the rural and urban communities. At the landing
sites fishers sell their catch to traders residing in that community or to “migrant” traders
who reside there for a brief period of time during periods of bountiful catch and lower
prices. These migrant traders may choose to either process the fish locally before it is
transported back to their area of residence or immediately transport it back to their home
base to be sold to processors there. The choice by the migrant trader would normally
depend on the relative prices of both freshly landed and processed fish, as well as cost of
processing and transporting the fish, which are all subject to change depending on market
University of Ghana http://ugspace.ug.edu.gh
14
conditions. Fish is mainly sold in its fresh or processed state in domestic markets located
all over the Ghana. Local supply of fish Ghana comes from the sea, lagoons, lakes, rivers,
aquaculture and imports. Due to the higher demands for fish in the southern and middle
portions of the country, fish marketing activities are most prominent in those areas (FAO,
2016) The capital city of Ghana, Accra is the most important local market and
consumption centre for fish and fish related products areas (FAO, 2016) Other important
consumption areas include Kumasi, Tarkwa, Tema and Sekondi-Takoradi.
Fish marketing is dominated by female fish traders popularly known as “fish mummies”.
These fish mummies normally fund fishing trips beforehand and buy fish directly from
those fishermen for onwards transportation to other actors in the fish value-chain (FAO,
2016). They also provide support in some cases, to processors by selling to them on credit
(FAO, 2016). As such in most cases they may be able to dictate or influence the prices of
fish both fresh and processed on the market. Fish mothers typically behave as a
monopsonistic (one buyer) cartel: they restrict access to fish supply through pre-
financing of fishing expedition (Gordon et al., 2011). It is this ability of fish mothers to
pre-finance fishing expeditions that give them unique access to fish landed for the day
and is also the reason why they can come together to affect the price of fish in the market.
From Gordon et al. (2011) processors of smoked fish sell their products in various
markets located in and around their district or region of origin. They normally sell their
products in packs to retailers who also sell to consumers in the same markets or in nearby
markets. Some traders also go directly to processors of fresh fish in the various localities
to purchase fish which they then sell to retailers in central markets. Most Processors have
a mutual agreement with each other whereby they sell on alternate market days so that
University of Ghana http://ugspace.ug.edu.gh
15
they do not flood markets with too much smoked fish. This would lead to a drop in the
prices of smoked fish and hence profits. Most if not all traders and processors for smoked
fish belong to an association for those involved in the smoking and trading of fish.
Members belonging to such an association will usually assist one another when the need
arises and also share information on the prevailing prices and supplies of processed fish in
the country. The head of the association is normally referred to as “commodity queens”.
They are responsible for implementing informal market rules and regulations as well as
resolving conflicts between association members.
In the lean season, when supply of fish is low, retailers that are well off purchase frozen
fish from cold stores in their various localities and process it themselves (Gordon et al.,
2011).
2.3 Processing of Fish
Due to the high water content in fish, it is an extremely perishable commodity. Fish
begins to spoil as soon as it is taken out of the due to the activities of microbes, which
result in an unpleasant taste and bad smell that deter consumers from purchasing it
(Obodai et al., 2009). The term fish processing refers to all the stages
fish and fish products pass through from the time of capture to the time it finds its way to
the final consumer. These methods include drying, frying, fermenting, smoking, salting,
and a combination of these methods to prolong the shelf life of fish. Refrigeration is also
another method for preserving fish. However, there are some bacteria that can survive in
cold temperatures and are only eliminated by the application of heat (Bender, 1982). In
Ghana; smoking is the most common method of fish preservation and processing.
Smoking preserves fish by drying and depositing natural wood smoke chemicals like
University of Ghana http://ugspace.ug.edu.gh
16
phenols and aldehydes on the flesh of fish. These chemicals preserve fish by preventing
the micro-organisms from surviving on the flesh of the fish (Garrow & James, 1994). It
was observed that about 80 percent of the fish captured in the sea ad from freshwater
sources is eaten in smoked form (Samey, 2015). Fish smoking in Ghana is traditionally
carried out by women in coastal towns and villages, along river banks and on the shores
of Lake Volta. Often want to keep some of the fish in storage for some months to take
advantage of a more favourable market. Fish Smoking is beneficial because it helps to
reduce post-harvest losses during seasons of bountiful fish harvest (July to September)
and also allows for the storage of fish against the lean season. Other benefits of food
include an improvement in the taste, increase in the availability of protein to people
throughout the year, ease in the packaging, transporting and marketing of fish.
2.4 The Concept of Markets
In the analysis of spatial price linkages, investigations on the spatial market integration of
agricultural markets can be used to determine if agricultural markets are performing their
functions efficiently. Oftentimes the integration of markets and the efficiency of markets
are sometimes used interchangeably as such a look at the concepts of markets, market
integration and market efficiency and the way in which they are related to one another is
very important. A market can be defined as a place where buyers and sellers of a
commodity interact with each other to facilitate the exchange of a commodity. Stigler
(1969) defined market as "the area within which the price of a good tends to uniformity,
allowance being made for transportation costs". A market involves all market actors
spread out to perform marketing activities. Market structure, market conduct and market
performance are important determinants of the extent of competition or marketing
University of Ghana http://ugspace.ug.edu.gh
17
efficiency. The degree of competition or efficiency of a market will usually influence the
degree of the market and hence the prices that will be occurring in that market.
The extent of efficiency of a marketing system depends on the structure of the market, the
market conduct and performance of the market. The concentration of sellers and buyers,
the size of firms and the conditions for entry into the market are the principle elements of
the structure of the market, influencing the degree of competition and pricing within the
marketplace. McCorriston et al. (1989) was one of the earliest researchers, to model the
structure of market as exerting some influence on the transmission of changes in prices.
In their study in 1989, they show that market power reduced the extent of the
transmission of price between the farm and retail stages. Thus, price transmission will be
completed if markets are imperfectly competitive in nature. Market conduct can serve as
an indicator of how agents in the market behave in relation to the determination of price,
tactics to promote sales and to regulations introduced by the government. The way in
which prices are formed is directly linked to these actions. When prices in the
marketplace are formed based on the collusive activities of market agents imperfect price
transmission will occur within that particular market or between that market and other
markets. This will result in inefficiencies in the marketing system as a whole. Prices
generated from the interactions of the forces of demand and supply is similar to that of a
marketing system believed to be efficient.
The structure of a market together with the market conduct is representative of the
performance of a market (Kanakaraj, 2010). Market performance is defined as how well a
market uses the scare resources available to meet the demand of consumers for goods and
services. In an economy, where a price determined by a firm is just equal to its average
University of Ghana http://ugspace.ug.edu.gh
18
cost, the market can be considered to be efficient or is performing it functions adequately.
In the same vein, a less profit margin can also be taken as an indication of market that is
efficient in performing its function. In other words, one can determine the efficiency of a
particular marketing system through the price and profit margin levels of prices and the
level of profit
2.5 The Concept of Market Integration, Price Transmission and Spatial Arbitrage
The existence of integration among markets is a useful tool in measuring marketing
efficiency in both temporal and spatial analysis. The focal point of a lot of studies on the
integration of markets market can be found in the Enke-Samuelsson-Takayama-Judge
(ESTJ) spatial equilibrium model (Enke, 1951; Samuelson, 1952; Takayama & Judge,
1971), in which the price differences for the same commodity in markets located at
different location is bounded from above by the cost of arbitrage between the markets
with no restrictions on the volume of trade and from below when volumes of trade
volumes reach a critical value (Abunyuwah, 2007).
Spatial integration of market can be defined as a situation where the price of a good in
markets located at different areas move together over time. Market integration is an
evaluation of the extent to which shocks to demand and supply in a particular market are
passed on to another market located in an area different from that of the other market
(Negassa et al., 2003). Barrett & Li (2002) defined the integration of markets in the
context of contestability and tradability between markets which involves processes that
clear the market, where supply, demand, and cost of transaction in different markets
influence the flow of trade and prices together, and the movements of shocks to price
between markets. Defining market integration in terms of tradability allows for flow in
University of Ghana http://ugspace.ug.edu.gh
19
trade to be a sufficient condition required to indicate spatial integration of markets but not
necessarily resulting in equalization of prices which is consistent with Pareto-inefficient
distribution (Barrett, 2005). That is to say two markets can be said to be integrated when
they by belong to the same network or when a government institution fixes prices to
allow them to adjust to shocks at the national and/or regional level making it possible for
price transmission to occur even when trade does not occur between the markets (Cirera
& Arndt, 2006). On The other hand, contestability looks at markets that are competitive
and the complete use of rents to arbitrage, as such two markets can be considered as being
integrated when there is the presence of zero marginal profits to arbitrage which makes
market participants unconcerned about trading thereby arriving at competitive
equilibrium and a Pareto-efficient distribution (Barrett & Li, 2002).
Spatial market integration is important in agriculture due to the bulky and perishable
nature of agricultural commodities; also the area of production of the produce and the
area of consumption may be different possibly implying expensive cost incurred in the
transporting the produce between locations (Sexton et al., 1991). Markets that are not
well integrated will often result in the transmission of information of price signals that are
incorrect to all participants along the marketing chain that will likely end in production
and marketing decision that do not benefit anyone (Goodwin & Schroeder, 1991). Also
markets that are not integrated do not allow economic agents to enjoy benefit derived
from of lower prices or increased productivity. This is because in markets that are not
integrated, the transmission of incentives of price and the positive impacts associated with
welfare are restricted (Ankamah-Yeboah, 2012)
University of Ghana http://ugspace.ug.edu.gh
20
According to Goletti et al. (1995) integration among market can be attributed to the
actions of traders as well as the environment in which they operate. These include
marketing infrastructure such as communication, transportation, storage facilities and
credit which generate large marketing margins due to the cost involved in the transfer of
goods. The actions of government such as policies aimed at stabilizing prices, restrictions
on trade and regulations on credit and transportation can affect the ways in which markets
function (Ankamah-Yeboah, 2012). Rapsomanikis et al. (2006) also identify the
oligopolistic and collusive behaviours among local traders as one of the determinant of
integration of markets; thus traders may maintain price differences between markets at a
level that is above that determined by transfer costs.
2.5.1 Price Transmission
Price transmission is thought to spring from three concepts (Balcombe & Morrison,
2002). The concepts are co-movement and completeness of adjustment, dynamics and
speed of adjustment and the asymmetry of response. The notion of co-movement and
completeness of adjustment implies that at any point in time variations in the prices of a
commodity in a particular market are completely passed on to another market. The
dynamics and speed of adjustment looks at the means by and speed at which, prices in a
particular market is completely passed on to another market or level. Finally, the
asymmetry of response component tries to identify if the price transmission process
among markets are characterized by symmetric or asymmetric relationships in other
words are shocks that increase prices passed on faster than those that decrease market
prices. The rate of adjustment and the degree of completeness can be characterized by
relationships that are asymmetric in nature.
University of Ghana http://ugspace.ug.edu.gh
21
From the above given notions, complete price transmission can only occur between two
spatially separated markets when variations in the price of a commodity in one market is
fully and instantly transmitted to the prices in the other markets for similar commodities,
as postulated by the Law of One Price (LOP). This implies that the markets for a
particular commodity are integrated. It also implies that when changes in prices are not
passed on immediately, but after a certain period of time has passed, then the process of
price transmission will be completed the long run and not the short run, as implicit in the
conditions of spatial arbitrage. The separation of price transmission into short run and
long run price component is very essential as is the rate of price adjustment to their long
run relationship. This is fundamental in understanding the degree of market integration in
the short run
2.5.2 The Concept of Spatial Arbitrage
Spatial Arbitrage is the process whereby traders transport commodities from one market
to another whenever the difference in price between the two markets is larger than the
cost involved in transporting a good between the markets. Spatial arbitrage can be
represented mathematically by formula:
i ij j
t t tp C P 2.1
i ij j
t t tp C P 2.2
𝑝𝑡𝑖 denotes the price of a commodity in the export market i in period t 𝑝𝑡
𝑗 denotes the
price of the same good in the import market j, and 𝐶𝑡𝑖𝑗
denotes the cost of transfers in the
same period. Equation 2.1 represent the condition for trade to occur and equation 2.1
University of Ghana http://ugspace.ug.edu.gh
22
represent a disincentive to trade. Equations 2.1 and 2.2 are known as the spatial arbitrage
conditions. When the variations in prices between two markets is greater than the cost of
transaction, opportunities for arbitrage are created and arbitrageurs make use of such
opportunities to make profit, by buying goods at lower price from surplus markets and
moving them to deficit markets where prices as higher. This is done until the variations in
price become equal to the transportation cost. If the difference in prices of a homogenous
commodity between a pair of markets is consistently above than the cost of
transportation, it is an indication that the markets are not working. This can be due to the
action of government such as placing restrictions on trade, collusive behaviour of traders,
lack of credit, lack of market information, or some other factors. The frequent occurrence
of spatial arbitrage opportunities in markets is an important indicator of market
efficiency.
The conditions of spatial arbitrage ensure that, in competitive markets that trade with one
another differences in the prices of commodities should be equal to the cost of
transaction, while at independence or self-sufficiency differences in prices between two
locations is either lower than or equal to the cost of transaction (Tomek & Robinson,
2003). The spatial arbitrage conditions also allow for the law of one prices to operate.
2.5.3 The law of One Price
The Law of One Price (LOP) follows directly from the spatial arbitrage condition. The
LOP states that “in the assumed absence of transport costs and trade restrictions, perfect
commodity arbitrage insures that each good is uniformly priced (in common currency
units) throughout the world” (Isard, 1977). For instance, Pjt and Pit are the respective
prices of a similar commodity traded between an export market j and an import market i
in the same period. The LOP (in its weak form) requires that the price differences
University of Ghana http://ugspace.ug.edu.gh
23
between i and j for the homogenous commodity be equal to the transfer costs involved in
transporting the commodity from market j to i. It stipulates that whenever the price
difference exceeds the transfer costs, arbitrage processes work out to ensure equality
between the inter-market price difference and transfer costs. The LOP mathematically
states that:
i j ij
t t t tp p c 2.3
Where, 𝐶𝑡𝑖𝑗
is the cost of transferring the commodity between market i and j whereas
𝜇𝑡 represents the deviations from the LOP in the short run due to immediate unexpected
shocks (e.g. failure of transportation systems, natural disasters, policy incoherence). The
equation (2.1) implies that:
i j ij
t t t tp p c 2.4
Earlier analyses of market integration that emphasized the concept of the LOP in the
above form mistook any inter-market price relationship that failed to fulfil the LOP as
market segmentation. The “law of one price” (LOP) can be used to determine the market
size, predict variations in price within a market and evaluate the efficiency in setting
prices of goods in a marketplace (Kohls & Uhl, 1998). The LOP will fail to operate in
regions not linked by arbitrage. Studies on the LOP by authors such as Baffles (1991) and
Ardeni (1989) found the LOP to be a phenomenon occurring in the short-run. They did
not find any evidence to indicate that the LOP was a long-run phenomenon. They
suggested that the failure of LOP to be a long run phenomenon is due to price,
institutional factors, cost of transaction and problems that are time–specific in nature.
University of Ghana http://ugspace.ug.edu.gh
24
This led to the modification of the concept of the LOP in co integration and regime-
switching models in which nonlinearities in price series are explicitly accounted for.
2.6 Conditions for Efficient Market Arbitrage
Conditions for efficient market arbitrage can be grouped into necessary and sufficient
conditions. The necessary conditions are satisfied when there is the exchange of
information and goods across space, time and form between any pair of entities engaged
in trade (Prajogo & Olhager, 2012). The sufficient conditions on the other hand are
satisfied when the variations in the price between two markets in different locations is
less than or equal to the cost of transaction (Prajogo & Olhager, 2012; Baulch, 1997).
That is to say that the LOP is working in these markets.
Efficient arbitrage in markets will lead to markets that are efficient in performing their
functions. When there is efficient arbitrage among spatially separated markets, higher
prices occurring in deficit markets will results in the flow of goods from surplus markets
to deficit markets (Negassa & Myers 2007). This will occur until prices in the two
markets become equal. This shows that the markets for that commodity are efficient.
2.7 Techniques for Measuring Spatial Market Integration
Studies on spatial market integration try to answer questions on the characteristics of the
price transmission process that occur among situated at different locations. This is done
by answering questions on the patterns of causality among market pairs, dynamic
interactions and the existence of a long-run equilibrium. Some of the methods used in
answering such questions are briefly discussed below.
University of Ghana http://ugspace.ug.edu.gh
25
Studies on market integration started with by using static price relationships as a test for
spatial integration among markets. This method involved estimating a bivariate
correlation and regression coefficients of similar commodities in different markets
(Hossain & Verbeke, 2010). This method was inspired by the idea prices in markets that
are integrated will tend to move together. According to this method a high coefficient of
correlation result was taken to imply market integration whereas a low correlation
coefficient implied that the markets were segmented. The static approach is a fairly
simple method; however, the model is plagued by weaknesses that can lead to the
drawing of wrong conclusions from results obtained. Cirera & Arndt, (2006) identify the
primary concern of this method be that correlation does not mean causality.
This procedure is not able to identify when the problem of heteroscedasticity is present in
the price data under study. Also where a lag in reaction to price change is caused by lags
in market information, a test of correlation will likely overestimate the absence of
integration among markets (Barrett, 1996). Analysis of bivariate correlation also hides the
masks the existence of factors such the effects of government policy effects and a general
increase in the level of prices (Goletti et al., 1995). The method also assumes that prices
adjust instantly to shocks and as such it is not able to capture the ability of prices in
markets to change. Results of price correlation coefficient are likely to misleading when
the data analysed is non stationary in nature (Wyeth, 1992). Finally, the static price
correlation method cannot be used to analyse the whole marketing system only pair wise
market analysis.
The Delgado Variance Decomposition Approach so named after the researcher Delgado
was developed in 1986 as an another model to measure market integration due to the
University of Ghana http://ugspace.ug.edu.gh
26
weaknesses of the bivariate model as mentioned above. According to Negassa et al.
(2003), this approach allows the whole marketing system to be tested for the existence of
market integration. This overcomes the shortcomings of the static price correlation
method. Before the introduction of this method, tests for integration among market
assumed that transaction and transport cost were constant and also allowed for the
removal of familiar trends and seasonalities present in price series. This then ensured that
the equality in spatial price spreads between markets pairs in a certain season could be
taken to mean that the markets are integrated. The shortcoming of this method is that it
does not allow for dynamic relationships among two markets in different locations. Also
the method is based on price relationships that occur in the same period.
The Ravallion approach developed in 1986 became the most popular method for
measuring integration among market. This was because of its ability to differentiate
between integration and segmentation of market in both the long and short run after
adjusting for seasonal changes, seasonality, trends and autocorrelation (Negassa et al.,
2003). Agricultural markets are slow to adjust when shocks that may call for substantial
time lags are introduced into the markets. This slow adjustment of agricultural markets to
such shocks is what informs the basis of the idea underlining this method. By linking
dynamic considerations in this model researchers are able to dodge the danger of
inference that is known to be present in the static model mentioned above. According to
Cirera & Arndt, (2006) the Ravallion model assumes that the cost of transfer among
market is constant and also neglects the possibility that there could be reversals in the
flow of goods between seasons. In situation where the cost of transfers is complex or vary
with time, any conclusion drawn will be prejudiced towards the inability to reject the
University of Ghana http://ugspace.ug.edu.gh
27
hypothesis of segmented markets. This method asserts that the nature of the structure of a
market existing between a number of domestic markets and a principal market is radial in
nature. This means that the prices formed in the domestic markets are controlled by trade
with the central market.
Regression technique has been used to examine integration among markets (Alexander
&Wyeth, 1994). The problem with this method was that regression analysis could only be
applied to stationary data as it application to data that as non-stationary will lead to
spurious regression. This shortcoming was resolved by using the first difference of the
price series in the regression, but this resulted in the loss of long-run information. Co-
integration techniques provided a way for analysing non stationary time series data that
did not produce spurious results. This helped to improve the strength of research findings
as price series used in testing market integration are often non-stationary in nature.
Cointegration was introduced by Engle & Granger (1987) and Engle & Yoo (1987). They
define cointegration as the presence of a long-run relation between various price series.
The nonexistence of a long run relationship between prices in two different market means
that the market pairs are segmented, while its presence indicates that the markets pairs are
integrated with each other (Ankamah-Yeboah, 2012).
Cointegration analysis begins by first finding out what the order of integration of
individual price series is by using a suitable unit root test. If all the individual price series
are integrated of the same order, a co-integration regression is created and residuals from
the regression are tested for the presence of unit roots. One of the weaknesses of the
cointegration method introduced by Engle & Granger is that it did not enable one to the
test of all possible co integrating vectors in a multivariate system. This then led to
University of Ghana http://ugspace.ug.edu.gh
28
Johansen (1988) coming up with a co integration approach which overcomes the
shortcoming stated above. The Johansen method relies on maximum likelihood in testing
for the existence of long run relationships among several economic series. Engle &
Granger (1987) also proposed to use error correction models in analysing the short run
dynamics if co integration was found to be present between the variables under study.
The error correction representation explains the process of adjustment of both short-run
and long-run reaction to variation in price which usually indicate arbitrage and efficiency
in markets (Abunyuwah, 2007). Co integration and error correction models can also be
used to explore ideas such as completeness, speed and asymmetry of price relationships
and the path in which causality between a pair of markets will occur. According to Barret
(1996) presence of long run relationship among prices is not a necessary or sufficient
condition for integration between markets. When transaction costs are not stationary, lack
of proof to support the presence of long run relations may be consistent with integration
of markets. Cointegration is not sufficient because a negative sign of the coefficient of the
price in the principal market implies divergence instead of co movement.
The Error correction model is an expansion of the co integration model. If there is the
presence of a long run relationship among two price series 𝑝𝑡𝑖 and 𝑝𝑡
𝑗 then the equilibrium
relationship between them can be defined as: 𝜀𝑡 = 𝑝𝑡𝑖 − 𝛽𝑝𝑡
𝑗− 𝛽0 and the error term 𝜀𝑡 is
presumed to follow an autoregressive (AR) process, then 𝜀𝑡 = 𝛼𝜀𝑡−1 + 𝑒𝑡 this shows that
the equilibrium relationship between 𝑝𝑡𝑖 and 𝑝𝑡
𝑗can be defined as:
0 1
i j
t t t tp p e 2.5
University of Ghana http://ugspace.ug.edu.gh
29
Equation 2.5 shows that co integration between 𝑝𝑡𝑖 and 𝑝𝑡
𝑗 is a function of the
autoregressive process 𝜀𝑡−1 where 𝜀𝑡−1 represents the deviation from equilibrium in the
long run and is known as the error correction term (ECT). The ECT measures the
response of 𝑝𝑡𝑖 and 𝑝𝑡
𝑗 to deviation from equilibrium. The standard ECM has been
expanded to asymmetric error correction model, vector error correction models (VECM)
and switching vector error correction models (SVECM).
Parity Bound Models (PBM) was developed by Spiller & Haung (1986) and Spiller &
Wood (1988). According to Abunyuwah (2007), the PBM was developed to try and use
all market data to assess the efficiency between markets located in different regions. The
model presumes that the cost of transaction plays an essential role in influencing the price
efficiency band or parity bounds (Barrett & Li, 2002). The PBM evaluates the degree of
integration among markets market integration by estimating the likelihood that an
observation will belong to one of the three regimes (no profitable trade, efficient trade,
and unexploited profitable trade opportunity). This approach is an improvement over the
above error correction models as it takes into consideration transaction costs, trade
reversals, and discontinuity (Barrett, 1996). The approach is, however, criticised for its
static nature and data limitation on observable transfer costs, especially in a developing
country setting
The threshold autoregressive models (TAR) was developed by Howell Tong, (1983). The
TAR assumes that the cost of transactions must go beyond a certain threshold before price
will begin to adjust towards equilibrium leading to integration of markets (Goodwin &
Piggot, 1999). The threshold effects will usually happen when shocks above a significant
threshold level result in responses that vary from those by shocks that lie below a
University of Ghana http://ugspace.ug.edu.gh
30
significant threshold. The thresholds are generally viewed as components of adjustment
and transaction costs or economic risks that keep market operators from constantly
adapting to variations occurring in markets (Rapsomanikis & Karfakis, 2007). Although
the ability of the TAR to identify the cost of transaction constraints in measuring
integration among markets market, the model still has some shortcomings. The limitation
of the PBM is that it that the costs of transactions are constant and as such implies an
unchanging neutral band over the study period (Abdulai, 2007). The addition of time
trends to the threshold and parameters of adjustment and the subsequent specification of
the threshold as a simple linear function of time is one of the attempts at trying to
overcome the weakness of the PBM (Van Campenhout, 2007). Another attempt
introduces sub-samples that are different in nature to show varying policy and economic
setting to capture possible changes in the cost of transaction as a result of different policy
regimes (Abdulai, 2007).
2.8 Asymmetry in price transmission
Asymmetry in price transmission exists when prices at different levels of a market
respond differently to shocks. According to Rapsomanikis, (2006), the presence of
asymmetries in the response of one price changes in another price shows that the
adjustment process is non-linear in nature. A lot of researchers have used the asymmetric
error correction model proposed by Granger & Lee (1989) or threshold co integration
models developed by Enders & Granger (1998) to study asymmetry in the price
transmission process.
Price asymmetry exist either in the speed or extent of adjustment or both. In terms of the
extent of adjustment, elasticities in the short-run vary according to the direction of the
University of Ghana http://ugspace.ug.edu.gh
31
initial change whereas with the speed of adjustment it is the long run elasticities that vary
(von Cramon-Taubadel, 1998). Asymmetry in price transmission can either be positive
or negative. Positive asymmetry occurs when prices respond more quickly to shocks that
increase price than to those that decrease prices. In other words, price adjustment that
squeezes the margin is transmitted faster those that widen the margin. Negative
asymmetry is when prices respond faster to a decline in another price than to an increase.
In other words, prices respond quickly to adjustment in price that will enlarge the margin.
This usually can be used to determine the path followed in transferring welfare (Meyer &
von Cramon-Taubadel, 2004). Vertical asymmetry is when asymmetry in price responses
occur along the food supply chain whereas spatial asymmetry is when asymmetry occurs
between markets located in different areas. APT means that markets actors are not
enjoying the benefits they would have derived from reductions or increase in prices under
conditions of symmetry. This is because under conditions of asymmetry the timing and/or
size of the changes in welfare linked with changes in price are distorted (Meyer & von
Cramon-Taubadel, 2004).
2.8.1 Causes of Asymmetry in Price Transmission
A possible cause of asymmetries in price transmission is market power. Market power is
when some actors in the market are able influence the prices of commodities such that it
remains over a certain competitive level (Amonde et al., 2009). In agriculture especially,
farmers and consumers often believe that imperfect competition in processing and
retailing provides an opportunity for middlemen to abuse market power. It is expected
that this will often result in positive asymmetry however; it is not the only effect
emanating from market power. Ward (1982) point out that because of the fear on the part
University of Ghana http://ugspace.ug.edu.gh
32
of oligopolists in losing market share, they may be hesitant to increase the price of their
output.
Adjustment costs are also considered as a cause of APT. Adjustment cost refers to the
cost incurred by firms in altering the input or output prices or quantities. When the costs
incurred are characterized by asymmetry in the increases or decreases associated with
quantities or prices. APT can occur. In the situation of changes in prices we can refer to
adjustment costs as menu costs (Meyer & von Cramon-Taubadel, 2004). Menu costs
include the cost incurred when nominal prices are changed, cost of inflation and the cost
of distribution information about changes in price. Asymmetric may be present in menu
cost with respect to rising and falling price level. When there is continuous increase in the
general price level menu cost can lead to the incidence of asymmetry (Ball & Mankiw,
1994). Under such conditions, Abdulai (2000) reported that firms respond more quickly
to shocks that increase prices since positive shocks help them to correct for expected
inflation.
Inventory management or stock behaviour of traders has been identified as another cause
of APT in a lot of markets. Instead of decreasing the price of output when demand is low
firms would rather increase their inventory level while in times when demand is high they
increase the prices of their goods. These combined with asymmetry in costs associated
with increases and decreases in the levels of inventory and the fear that stock will run out
may lead to prices responding rapidly to increases than to a decrease (Reagan &
Weitzman, 1982). Balke et al. (1998) also point out that methods of accounting such as
FIFO (first in first out) can lead to APT.
University of Ghana http://ugspace.ug.edu.gh
33
Also the role of government can bring about asymmetry in price transmission. It is
common in agriculture for governments to intervene by providing price support such as
price floors. This is because such mediations by government can lead to APT if
wholesalers and retailers are convinced that the decrease in the prices will not last for
long especially since it will bring about government intervention, whereas increases in the
prices is likely to have a lasting effect.
2.9 Empirical Studies linked to Market Integration Price Transmission in Ghana
and Elsewhere
Ghanaian agricultural markets have been subjected to a lot of research on the behaviour
of price and their responses to one another most especially in the maize market with very
few studies delving into other markets such as the tomato, plantain and cassava.
In analysing price transmission between major maize markets in Ghana from 1980 to
1997, Abdulai (2000) utilized the threshold cointegration model. Results indicated that
maize prices between the time periods in question in the Accra and Bolgatanga markets
react faster to increases in the principal market price than to decreases. Abdulai (2000)
also found out those markets in Accra react faster to changes in prices in the Techiman
market prices than do markets in Bolgatanga.
Cudjoe et al. (2008) studied the effect of crisis on a global level on transmission of food
prices and poverty in Ghana. The results of the correlation analysis found out that
correlation between the price of rice imported into the country and the price of domestic
staples such as cassava, maize, rice and yam high in the Tamale and Wa markets found in
the northern region of the country. Coefficient of correlation was also found to be high
between the global prices of rice and maize. Cudjoe et al also applied the Johansen co
University of Ghana http://ugspace.ug.edu.gh
34
integration test which revealed price transmissions in Ghana where heterogeneous in
nature. No evidence of price transmission was found for root crops such as cassava and
yam across different regional markets, however in the market for cereals the transmission
of price was high in both the short and long run.
Amikuzuno (2010) employed TAR model to examine price transmission between four
major markets for fresh tomato in Ghana. The results of the study revealed that while
transmission of changes in prices improved for some pairs of market, while for other pairs
it deteriorated from periods of high to low tariffs. The results also showed an
improvement in the price adjustment, which assesses the price transmission level, and
thresholds that represent the cost of transaction after the reduction in import tariffs.
According to the author, other factors that can explain the observed signals of failure in
markets need to be examined. Also options to help reduce volatility in prices and also
enhance the competitiveness of fresh tomato in Ghana need to be considered.
Mensah – Bonsu et al. (2011) examined whether the plantain marketing system in Ghana
between the periods 2004 to 2009 were efficient. The authors used the Johansen test for
cointegration and error correction models to examine if plantain markets were integrated
or not. The Accra market was used as the main consumption market; markets in
Koforidua, Kumasi and Sunyani as assembling markets and finally the markets in Begoro,
Goaso and Obogo were selected as producing markets. Results of the study identified
short and long run relationship between markets studied implying that arbitrage works in
the plantain marketing system. Results also show that the rate at which prices were
transmitted between the Accra market and the other markets was 27.7% which was
University of Ghana http://ugspace.ug.edu.gh
35
moderately weak when likened to an ideal adjustment of 100% threshold. This result
suggested that further integration especially in the short run is required.
Acquah et al. (2012), applied the Johansen co-integration and Granger Causality test and
Autoregressive distributive lag model (ARDL) to examine the transmission of prices and
integration of local cassava markets in Ghana. The study used weekly prices of cassava
from August 2008 to April 2010. The results showed that although the price series were
integrated of the same order; long run relationship among the price series studied was
non-existent. The markets pairs under study were also found to be independent of each
other. This suggested that transmission of price and integration of cassava markets under
study was poor. The authors therefore recommended that the factors that influence spatial
price transmission should be critically looked at.
Acquah & Owusu (2012) also examined spatial integration of market and transmission of
prices of some markets for plantain in Ghana. Using co integration analyses and vector
error correction modelling to provide an econometric analysis of plantain markets
integration, the authors tested the degree of spatial market integration using prices from
three markets namely Techiman, Accra and Kumasi. The test of co integration identified
at least three co integrating relationships which showed that markets for plantain were
joined together. The result of the VECM revealed that a shock introduced to the market in
Techiman will be erased at rate 47% in the following period. This result suggested that
the rate of adjustment of price to shocks between the Techiman market and the other
markets was moderate given an ideal adjustment threshold of 100%.
University of Ghana http://ugspace.ug.edu.gh
36
Since most price transmission and market integration analysis in Ghana have focused
more on grains such as maize and rice rather than on other essential commodity like fish
which also plays an important role in the economic development of Ghana we look at
market integration and price transmission studies undertaken outside Ghana most
especially on the fishing sector.
Norman-López & Asche (2008) in studying the competition between tilapia imported into
the United States (US) and domestic catfish market using the Johansen cointegration test
found out that there was only one market in which for domestic catfish, fresh and frozen
catfish fillets were sold for the period (1997–2006) under study. Results also showed that
there were different markets for fresh and frozen tilapia fillets. No form of competition
was found to exist between the fresh and frozen catfish and tilapia products. A study by
Norman-López and Bjørndal in the following year confirms these results, by revealing
that the tilapia products imported into the US from Africa, Asia, Central and South
America were not co integrated with the US for the period 2002–2006. Also, no
relationship was found to exist between whole frozen tilapia and frozen tilapia fillets in
markets in the US.
Shinoj et al. (2008) studied the extent of price transmission and integration between the
major coastal markets in India. The authors used monthly retail price data on important
marine fish species for a ten-year period from January 1998 to December 2007. Markets
in the Andhra Pradesh, Gujarat, Karnataka, Kerala, and Maharashtra, Orissa, Tamil Nadu
and West Bengal states were chosen for the study. The fish species studied included
Mackerel, Sardine, Pomfret and shrimp. Results revealed that the Mackerel market had
the highest integration among the markets under study with the Shrimp markets appearing
University of Ghana http://ugspace.ug.edu.gh
37
to have the least market integration. Authors also discovered that the transmission of
prices between the markets in Kerala and Tamil Nadu was nearly complete except for
shrimp. Also the price movement in the market in Maharashtra was found to be
independent of the other markets despite its status as a major landing centre in India. The
study therefore suggested that strategies that will increase market integration be
introduced so that producer and consumer can benefit from it.
In Nigeria, Bada & Rahji (2010) in studying the market delineation of market for fish in
Nigeria for the period 1970–2005 using the Johansen Cointegration test showed that
prices of locally produced North African catfish in Nigeria is influenced by Hake,
Mackerel and Sardinella imports but not vice versa. The analysis also reveals that Catfish
could be used as a substitute for fish imported into Nigeria. All the species were found to
be in the same market and could be used as substitutes for each other. It was
recommended that the effect of imported fish prices needs to be taken into account when
designing fish production policies.
Adenegan & Bolarinwa (2010), examined the extent of price transmission and
integration of markets local fresh and dried fish and imported iced and dried fish markets
in the urban and area of Oyo State in Nigeria. The authors employed the use of the ADF
test, test for Granger-causality and Market Concentration index to analyse monthly retail
prices of fish for a period of five years. Data was obtained from the Oyo State
Agricultural Development Project the result of the study revealed that all market pairs
were well integrated except the market for rural and urban local fresh fish. This showed
that prices of fresh fish in urban and local market did not move together in the long run.
The granger causality test carried out for fifty-six markets pairs show that out twenty-five
University of Ghana http://ugspace.ug.edu.gh
38
market pairs did not granger cause each other, seventeen market pairs showed
unidirectional causality and the remaining market pairs showed bi-directional causality.
The market occupying the leadership position in forming price was found to be the urban
fresh fish market. The market connection index showed that the markets studied showed
that integration of markets was low in the short run. It was suggested that policies aimed
at improving efficiency of fresh fish markets in the urban and rural area of the Oyo state
be implemented.
Mafimisebi (2012) using a vector autoregressive approach looked at the spatial
equilibrium, integration of markets and exogeneity in price for dry fish in Nigeria. Retail
data on the monthly price of dried fish from January 1997 to December 2008 for twelve
state capital markets was used in the study. The results of the study disclose that all the
price series under study were integrated of the same order. The results of the VAR
revealed that more than half of the market studied integrated in the long run. The markets
in Akure, Bauchi, Kano and Makurdi were found to influence the prices in the other
market location. Only the markets in Kano exhibited Kano exhibited very strong
exogeneity. The study concluded that there the degree of spatial pricing efficiency in dry
fish markets in Nigeria was low. Thus it was recommended that market infrastructures
and the collection, collation and dissemination of information be enhanced. The author
also recommended that policy reforms that target a reduction in the price at the retail level
for identified leader markets be implemented, as a means to increasing spatial pricing
efficiency in dry fish markets in Nigeria.
Sapkota et al. (2012), their paper titled “Price Transmission relationships along the
seafood value chains in Bangladesh: aquaculture and capture fisheries” investigate the
University of Ghana http://ugspace.ug.edu.gh
39
causal and price transmission relationships existing between and retail prices of Catla,
Hilsa, Pangas Rohu and Tilapia fish species in Bangladesh from October, 2005 to July,
2010 for Chittagong, Dhaka, Khulna and Rajshahiu. The study employed the use of the
Granger causality and the Houck and Ward approach as well as the von Cramon-
Taubadel and Loy error-correction model for analysis. The results show in many of the
value chains analysed retail prices were found to influence prices in the Bangladesh fish
sector. The results also revealed in the short run; transmission of prices was characterized
by symmetry while in the long run a mixture of symmetric and asymmetric relationships
was observed. Variations in the transmission of prices between capture fisheries and
aquaculture products were also identified from the results of the study.
Bukenya & Ssebisubi (2014) used the Johansen Cointegration test to analyse monthly
price data from 2006 to 2013 in their study titled “Price integration in the farmed and wild
fish markets in Uganda” revealed that wild harvested and locally farmed North African
catfish in Uganda belong to the same market. Gordon & Ssebisubi (2015) in their paper
titled “Vertical and horizontal integration in the Uganda fish supply chain” also produced
a result similar to that of Bukenya & Ssebisubi (2014). They applied the Johansen
cointegration test to monthly price data from Uganda for the period 2006–2010. The
authors demonstrated that locally produced and wild-caught North African catfish in
Uganda belong to the same market together with farmed and wild-caught Nile tilapia and
wild-caught Bagrus. However, it was found out that Mukene and Nile perch caught in the
wild did not follow the same trends.
University of Ghana http://ugspace.ug.edu.gh
40
2.8 Summary
Observations made from the reviewed literature suggest that agents in the market are able
to make timely decisions when market information is easily accessible. To assess
integration of markets and price transmission of various agricultural commodities
numerous analysts have used various methodologies. Based on the law of one price, co-
integration suggests that if two markets are considered to be integrated, then a price
changes in one of the market is transmitted on a one for-one basis to the other market in
the short run, or over time in the long run. To test the assumption, several techniques
were identified in the literature, including co-integration, causality; symmetry; and an
error correction mechanism. The framework for this study combined several techniques to
better understand the degree of market integration. The notion of cointegration, which
accommodates both the short and long run responses, has not yet been applied to the
study of integration of processed fish markets in Ghana
University of Ghana http://ugspace.ug.edu.gh
41
CHAPTER THREE
METHODOLOGY
3.0 Introduction
This chapter deals with the models and procedures employed in carrying out this study. It
discusses the basic theory underlying price transmission and market integration analysis.
It also presents the empirical model employed in achieving the objectives set out in the
study, the source of data for the study and a description of the study areas.
3.1 Theoretical Framework
The theory behind integration of markets and price transmission analysis lies in
understanding how prices among markets in different locations interact with each other.
Understanding how prices interact in markets situated at different location that is the
spatial integration model, predicts that the difference in price for a pair of markets for a
homogeneous commodity under competitive conditions will approximately equal the cost
incurred in moving the homogenous commodity from one market to the other. The law of
one price as explained in chapter two is of two forms, the strong which says that the
difference in prices should equals the cost of moving the homogenous commodity from
one location to another and the weak places the price difference to vary by at most the
cost of transportation. The weak form of LOP is known as spatial arbitrage.
ij
i jp p C 3.1
According to Chirwa (2000), LOP assumes that changes in prices in markets that are
integrated will be transferred instantly on a one-for-one basis to among them. The
University of Ghana http://ugspace.ug.edu.gh
42
principle of spatial integration predicts that under competitive conditions, differences in
prices for a pair of market for a homogenous commodity will be nearly equal to the cost
of transactions. This is a long run concept; however, prices can move away from each
other in due to a variety of shocks in the short run. Price signals will stimulate the price
transfer of commodities between excess and surplus markets if a situation of
disequilibrium arises as such equilibrium will then be established in the long run. The
existence of a situation like this suggests a stationary term which can be defined as the
temporal and stochastic deviation from equilibrium. A common feature of stationary
series is that they constantly return to and fluctuate around their mean. This feature can be
referred to as a long run affinity (Rico, 2009). This in simple terms means that provided
the series is stationary, values like it mean and variance may change but as times goes on
will revert back to their average values. Such conducts are closely related to the economic
understanding of equilibrium which is also a long run concept.
Based on this theory it is expected that prices in surplus (producer) and deficit (consumer)
markets will be equalized in the long run even if there are deviations in the short run.
3.2 Conceptual Framework
Under competitive conditions, price differences between two regions in the same
economic market environment for a homogeneous commodity should approximately
equal to the inter-regional transportation costs. Market integration thus involves a test of
price efficiency by examining how food markets in different regions respond jointly to
supply and demand forces. If price movements in different parts of the country tend to
behave similarly, reflecting the cost of transferring the product between two regions, then
markets are said to be integrated. The price movement in the different markets provides an
University of Ghana http://ugspace.ug.edu.gh
43
important indicator as to whether the markets are integrated or not. If the prices in the different
markets for the same commodity move in identical patterns, those markets can be considered to
be well integrated.
Since market integration is important in indicating transmission of price signals and
shocks among commodities over time, the model underlying market integration
postulates, that there exist linkages among markets and stable relation among prices in
different localities. If two markets, a surplus market (𝑥) and a deficit market (𝑦) are
integrated then supply and demand shocks will cause food to flow from the surplus
market 𝑥 to deficit market y where prices are higher, thereby decreasing the food supply
in market x. The prices in market y would go down because of the increased supply from
market x and this would eventually lead to the prices in market x also going up due to the
decreasing food supply in market X. This co-movement of prices gives a degree of
market integration among the markets.
3.3 Empirical Framework
The definition of price transmission provided in the chapter two earlier contains within its
meaning the situation of perfect integration among markets, the inbuilt dynamic
interactions of market that crop up due to inactivity or interruptions in trade, as well as
nonlinearities that come to light as a result of policies and other distortions in arbitrage.
Even more important, is the identification of hypotheses, through its various elements,
that can be tested within a co integration–error correction model framework. A variety of
time series procedures exist to enable one assess each of the mechanisms involved in the
transmission of prices and eventually assess the degree of transmission of price changes
University of Ghana http://ugspace.ug.edu.gh
44
and integration between these markets. These procedures are depicted in the diagram
below.
Figure 3.1 Empirical framework for assessing price transmission
Source: Adopted from Rapsomanikis et al. (2006)
From figure 3.1, each price series is first tested to determine its order of integration using
the Augmented Dickey-Fuller (Dickey & Fuller, 1979) or the Phillips & Perron tests
University of Ghana http://ugspace.ug.edu.gh
45
(Phillips & Perron, 1988). If the price series are found to have been integrated of varying
orders, the study concludes that the markets are not integrated and carry out a test for
granger causality. If it is determined that the price series are stationary or integrated of
order zero I (0), the study assesses the dynamics of the price relationship that exist
between the series by means of Autoregressive Distributed Lag (ARDL) model and then
proceeds to test for Granger Causality.
In the case where it is determined that the price series were not stationary or integrated of
the same order that is higher than zero, the study test for the null of no cointegration in
contrast to the alternative hypothesis of at least one cointegration relationship using the
Johansen approach (Johansen 1988, 1991). The rejection of the null hypothesis indicates
co-movement of prices and that markets under study are integrated. If the study is unable
to obtain evidence against the null hypothesis, the study concludes that the markets are
not integrated and/or that that transmission of prices along the supply chain may be
incomplete. An ARDL model can then be used to analyse the dynamics of price relation
in such series.
After rejection of the null of no cointegration among price series a Vector error
correction model is used to analyse the dynamics of prices in short and long run, the
adjustment speed and the direction of long or short run Granger causality.
Finally based on the results on causality direction, an Asymmetric error correction model
is specified and tests the null of symmetry against the alternate of asymmetry. After
which we discuss the results and touch on the characteristics of transmission of price
changes and shocks and integration of market.
University of Ghana http://ugspace.ug.edu.gh
46
The frame work described above does not show the factors responsible for the integration
of markets and transmission of prices. That is the study is unable to point out whether
price transmission and integration of markets is influenced by the cost of transaction,
policies undertaken to insulate local markets, or the extent of market power wielded by
participants in the supply chain.
3.4 Method of Analysis
This section presents the methods used to analyse the data for the study.
3.4.1Analyzing the extent of market integration
To analyse the extent of market integration among markets, a unit root test is first
performed using the Augmented Dickey-fuller test. Studies on integration of markets first
begin with testing for unit root in price variables in order to determine if the co-
integration approach will be an appropriate tool the appropriate tool (Fossati et al., 2007).
The number of lags to be in the unit root test is chosen to ensure the absence of serial
correlation using suitable information criterion test. Lovendal et al. (2007), recommends
that, before the estimating a model, the existence of unit roots must be examined using
Dickey-Fuller tests to test for stationarity the data sets in order to avoid the issue of
spurious regressions and its related problems. The unit root test is performed first at level
then at first difference to check if the price variables are stationary or not and also their
degree of integration.
Mathematically stationarity can be shown as:
1t t tP P 3.2
University of Ghana http://ugspace.ug.edu.gh
47
where 𝜇𝑡 is a random walk with its mean, variance and co variance being equal to zero.
The process tests the null of presence of unit root, that is, 𝐻0: ∅ = 1 against the alternate
hypothesis of stationary, that is, 𝐻𝑎: ∅ < 1. Rejecting the null hypothesis implies that the
time series data is stationary.
After determining that their price series do not possess unit roots, we then move on to
specify the Johansen’s co integration to find out if the system of equations are co
integrated or not.
3.4.1.1 Johansen Cointegration
Co integration can exist between two or more variables if each variable on its own is non
stationary but a linear combination of the variables may be stationary (Gopal et al., 2009).
If two prices are found to be cointegrated it means prices in the long will follow the same
path or move together. The presence of Cointegration also show that granger causality
must exist in at least one direction among the variables and also that the co integrated
variables can be expressed in an error correction form. Cointegration can also be thought
of as a special case where a regression of series that are not stationary does not result in
spurious regression results.
The first step to consider before carrying out a cointegration analysis is to determine the
order of integration of the variables since this is an important because Cointegration can
only be carried out series with the same order of integration or in other words non
stationary. The order of integration is specified by the number of times a series needs to
be differenced in order for it to become stationary, it can be written as I (d) where d
represents the number of times the series was differentiated before becoming stationary.
University of Ghana http://ugspace.ug.edu.gh
48
The Johansen Full Information Maximum Likelihood test by Johansen & Juselius (1990)
is preferred to that proposed by Engle & Granger (1987). The test proposed by Engle &
Granger (1987) has the major disadvantage of assuming a single co integrating vector for
the regression. As such in the likelihood of there being more than one co integrating
vector this method becomes unsuitable for co integration analysis and the Johansen test
becomes the more suitable option. This study therefore adopted the Johansen Full
Information Maximum Likelihood test for its co integration analysis due this advantage.
The Johansen maximum likelihood approach is based on the vector auto-regressive
(VAR) approach to cointegration which assumes all variables are endogenous. The test
tells as what and how many co integrating vectors are in a set of I (1) series
Johansen & Juselius (1990) propose two methods to estimate the number of cointegrating
vector namely the Maximum Eigenvalue test which tests the null of r co-integrating
vectors against the alternative of r+1 co integrating vectors. The test statistics are
computed as:
max 12 ( : 1) (1 )rLn Q r TLn 3.3
The Trace test which examines the null of r co integrating vectors against the alternative
of r >1 co-integrating vectors, the trace test statistics is specified as:
1 12 (1 )
p
trace irLnQ T Ln
3.4
The trace test carries out a joint test for the eigenvalues whereas the maximum eigenvalue
test carries out individual tests for the eigenvalue. The Trace and Maximum eigenvalue
test most of the time provides similar results, however in the situation where they differ,
University of Ghana http://ugspace.ug.edu.gh
49
the results of trace test are chosen over those of the maximum eigenvalue test (Alexander,
2001). If a long-run equilibrium relationship is found to exist between time series data,
then a vector error correction model (VECM) is used to assess the features of the co-
integrated series in the short run. A VAR or ADRL model can be employed to evaluate
the price dynamics in the short run if the time series data is not co-integrated. Moreover,
the Johansen test also allows for the testing of restrictions on the cointegration relations β
and the adjustment speeds α in the VEC model.
The Cointegration approaches proposed by both Johansen and Engle and Granger assume
the adjustment mechanism is linear and symmetric and as such in the presence of
asymmetric adjustment these test for cointegration and their augmentations become
incorrectly specified and have low power (Enders & Siklos, 2001; Enders & Granger,
1998).
3.4.1.2 Test for Granger Causality
Granger causality is important in showing the direction of the causal relationship after
performing co-integration tests. A processed fish market price series can be said to
granger cause another processed fish market series if the present and past values of the
former improve the prediction of the latter price series. Rashid (2004) argues that
causality measures the ability of prices to be predicted i.e. the movement of price in one
market can be used to predict changes in price in other markets which can be tested
within Johansen’s co-integration framework.
Granger causality in markets can be manifested in three major ways i.e. unidirectional,
bidirectional or as independent price series (Gordon et al., 2011). Unidirectional
University of Ghana http://ugspace.ug.edu.gh
50
represents those markets in which shocks in one market cause prices in the other market
but there is no reverse or feedback effect. The market which granger-causes the other in a
unidirectional causality situation can be referred to as the exogenous market (Mafimisebi,
2012). Bidirectional causality on the other hand implies that shocks in a particular market
are passed on to another market and vice versa. In the bidirectional causality, prices are
said to be determined by a Simultaneous Feed Back Mechanism (SFM) that is there is
feedback from mechanism operating between the markets (Mafimisebi, 2012). However,
in a situation where none of the markets is granger causing the others there is independent
causality. In this scenario shocks in a particular market are not transmitted to other
markets unless under special conditions. Granger causality was conducted in this study to
determine which market caused the other. The Granger causality models applied in this
study are specified in the equations below:
1 1 1
1 1
n ni j i
t k t k t t
k k
P e P f P
3.4
1 1 2
1 1
n nj i j
t k t k t t
k k
P g P h P
3.5
Where 𝑃𝑖 is the consumer market and 𝑃𝑗 the producer market. The equations (3.4-3.5)
postulate that 𝑃𝑖 is dependent on 𝑃𝑡−1𝑗
and 𝑃𝑡−1𝑖 ; while 𝑃𝑗 is also dependent on
𝑃𝑡−1𝑗
and 𝑃𝑡−1𝑖 . The error terms are assumed to be uncorrelated error. For equation 3.4
and 3.5 rejection of the null hypothesis by a suitable F-test indicates that prices in the
market on the RHS granger-cause prices in markets on the LHS which implies that past
University of Ghana http://ugspace.ug.edu.gh
51
values of the series on the right hand side are adding information on the actual values of
the series on the left hand side, in addition to what is provided by its own past values.
Before Granger causality is carried the it is important to first determine the number of
lags to be used in the test as the direction of causality may be depend critically on the
number of lags (Gujarati, 2008). The granger causality test assumes that the variables are
stationary and that the error term is uncorrelated.
This Granger test in a price transmission analysis is important because it allows us to
understand which of the two prices acts as a source of information for the other and
enables us a gain qualitative understanding of the results, in terms of the causality
direction and the extent of market integration.
3.5 Extent of Price Transmission
The extent of price transmission among the markets was assessed using the Asymmetric
vector error correction model and the Autoregressive distributive lag model.
3.5.1 Extent of Price Transmission in Processed Fish Market (ARDL)
In running a regression involving time series data, if the regression model includes the
current and lagged (past) values of the explanatory variables, it is referred to as a
distributed-lag model. On the other hand, if the model includes one or more lagged values
of the dependent variable as explanatory variables in addition to the current and lagged
(past) values of the explanatory variable, it is called an autoregressive model (Gujarati,
2008). This model runs a regression with the explanatory variables and their lagged
values as well as the lagged values of the dependent variable on the left hand side of the
equation. In integration of market analysis, the ARDL is used as alternative test if we do
University of Ghana http://ugspace.ug.edu.gh
52
not reject the null of no cointegration. Since the price series were of a unit root and were
not integrated, their first difference was used to run the regression of the form:
1 1 1 t t t t tY X X Y u
3.6
Where Yt is the current price of the reference market (Makola), Xt is the current price in
the other market and Yt-1 and Xt-1 are the lagged prices in the reference and other market
respectively.
3.5.2 Extent of Price Transmission in Processed Fish Market (VECM)
The Vector Error Correction Model (VECM) was used to analyse the extent of Price
Transmission in the processed fish Market. VECM is a suitable model for variables found
to have one or more co integrating relationship vectors since it adjusts to both Short run
and Long run. The VECM is applied to measure adjustments in price induced by
deviations from the long-term equilibrium (ECT). The Error correction term (ECT) is
assumed to be a continuous and linear function of the magnitude of the deviation from
long-term equilibrium. As such any small deviations from the long-term equilibrium will
always lead to an adjustment process in each of the market. The VECM model is
embedded in the Asymmetric vector error model (AVECM), this is because in the
AVECM the ECT of the VECM is spilt into a positive and negative component. Due to
this the AVECM was used in analysing the extent of market integration and asymmetry of
cointegrated price series.
University of Ghana http://ugspace.ug.edu.gh
53
3.6 Estimation of Extent of Asymmetry in processed fish Market
The asymmetric error correction model is used to analyse the extent of asymmetry in
processed fish markets. Early studies on asymmetric price transmission a variant of the
variable splitting technique developed by Wolffram (1971) and later improved on by
Houck (1977) and Ward (1982).The technique involves splitting a variable Xt into a
positive component such that 𝑋𝑡+ = 𝑋𝑡 for all 𝑋𝑡 ≥ 0 and a negative components such
that 𝑋𝑡− = 𝑋𝑡 for all 𝑋𝑡 ≤ 0. The variable splitting technique refined by Houck (1977)
has been used in Market integration and price transmission analysis in an effort to explain
asymmetric adjustments. In this framework, the response of a price P1 and to another
price P2 is estimated with the following equation:
1 0 2 2
1 1 1
T
t t t t
t t t
P P P
3.7
Where ∆𝑃+ and ∆𝑃− P represent the positive and negative changes in P respectively.
𝛽0, 𝛽+ and 𝛽− are coefficients and T is the current time period. Asymmetry is tested in
the model by determining whether 𝛽+ = 𝛽−. Some analyst introduce long-run term in
∆𝑃2𝑡+ and ∆𝑃2𝑡
− to distinguish between short-run and long-run asymmetry. Long run
symmetry is tested by determining if the sum of the co-efficient in the polynomials is
equal, whereas short-run symmetry is tested by determining if the polynomials are
identical. Von Cramon–Taubadel & Loy (1996) provide evidence to show that the model
is basically not compatible with Cointegration between two price series. Granger & Lee
(1989) extended the Error Correction Model specification to allow for asymmetric
adjustments by splitting the ECT into positive and negative components. The resulting
Asymmetric vector error correction model is
University of Ghana http://ugspace.ug.edu.gh
54
1 0 1 1, 1 2, 1 2 1 2 1
1 1
n n
t t t t t t
i i
P P P ECT ECT
3.8
2 1 1 1, 1 2, 1 2 1 2 1
1 1
n n
t t t t t t
i i
P P P ECT ECT
3.9
Where 𝜀𝑡 … . 𝑁(0, 𝜎2), 𝐸𝐶𝑇𝑡−1+ and 𝐸𝐶𝑇𝑡−1
− measure adjustments to positive and negative
shocks respectively. Since 𝐸𝐶𝑇1+ + 𝐸𝐶𝑇1
− = 𝐸𝐶𝑇𝑡 we can thus see that the standard
symmetric VECM is embedded in the AVECM and the F-test to test the null hypothesis
of symmetry is given by 𝐻0 = 𝛽2+ = 𝛽2
−. If this is rejected, then price transmission
process among the market is asymmetric. This extended model of AVECM was used for
this study and in this case the variables being modelled are prices at different markets that
are spatially related which imply that a positive (negative) ECT indicates that the
marketing margin is above (below) its long run equilibrium. The idea that increases in
producer price are passed on faster than decreases in producer price can then be
formulated into a testable hypothesis that positive ECT values are corrected more rapidly
than negative ECT values.
The speed of adjustment parameters can also be expressed as a half-life Thalf, which
shows how long it will take for half of the deviation from long-run equilibrium to be
corrected. The half-life is computed using the following formula: 𝑇ℎ𝑎𝑙𝑓 = 𝐿𝑛(0.5)/
𝐿𝑛(1 + 𝛽), where 𝛽 is the adjustment parameter estimate from the AVECM. In all the
estimates, since the time of observations is monthly, the values of 𝑇ℎ𝑎𝑙𝑓 are calculated in
months.
University of Ghana http://ugspace.ug.edu.gh
55
3.7 Study Area and Data Source
Secondary data on the price information of koobi, smoked herrings and Kpala from the
markets being considered were used in the study. The secondary data used in the analysis
is presented in appendix 1, 2 and 3. Ten processed fish markets namely Ada, Agbozume,
Cape Coast, Half-Assini, Kpandu, Makola, and Mankessim. Techiman, Tema and Wa
were selected for the study. The locations were selected based on the availability of data
and geographical location of markets. Most markets were located close to the sea with a
few of them being far from the coast. This was done to see how prices are transmitted
among markets close to and far from the source of fish (sea) and if such markets were
well integrated with one another.
The markets are located in Greater Accra Region, Upper West Region, Central Region,
Western Region, Volta Region and Brong-Ahafo Region. Among the Regions, Greater
Accra is the most important domestic fish market and consumption centre in Ghana
(Fishery and Aquaculture Country Profile. Ghana. 2016). The Greater Accra region has
the Makola market as one of its most important market. Hence, the Makola market is
taken as the principal market. The data was obtained from the Ghana Statistical Service
(GSS). The study used price information from January 2012 to December 2017, a total of
72 observations.
Markets that were missing two or more consecutive and recurring price values were
dropped especially in the case of the Kpala markets. The unit of measurement for the
price data collected from GSS were for GH¢/kg of koobi, smoked herrings, and Kpala.
Log transformations of all the price series was carried out to make it easier to interpret
parameters and also because of the chance to reduce the heteroscedasticity in the price
University of Ghana http://ugspace.ug.edu.gh
56
information. Since the prices were compared at the same market level and for the same
commodities deflation of the price series was not done.
Figure 3.2: Map of Ghana showing the locations of the market under study
Source: Ezilon Maps (http://www.ezilon.com/).
University of Ghana http://ugspace.ug.edu.gh
57
CHARPTER FOUR
RESULTS AND DISCUSSIONS
4.0 Introduction
Discussion of the results emanating from the study is presented in this chapter. The
discussion on the variability and movement of processed fish prices is presented in
section 4.1. The stationarity analysis of the time series on the price data and the test of
cointegration are discussed in section 4.2 and Section 4.3 respectively. Section 4.4
presents discussions on results of Granger causality test. Section 4.5 present discussions
on the short- run and long-run dynamic interrelationship between the processed fish
market pairs. The results of asymmetry in price transmission in the processed fish market
pairs are discussed in section 4.6
4.1 Descriptive Analysis of Processed Fish Markets
Changes observed in agricultural prices at different points in time may often be attributed
to fluctuations in yield, production, seasonality, condition of infrastructure, government
policies and the behaviours of consumer and other market participants. It is imperative
that we have a clear idea of the nature of variability in agricultural prices over time and
across space and the sources of these variations before we analyze price linkages
(Ankamah-Yeboah, 2012). Table 4.1shows summary statistics of processed fish prices in
the study areas.
4.1.1 Descriptive Statistics of Processed Fish Prices
Across the markets, the highest maximum price and highest average price of Koobi was
observed in the Makola market at GH₵19.60/Kg and GH₵15.28/kg respectively while
both the lowest minimum price and lowest average price of Koobi was observed in the
University of Ghana http://ugspace.ug.edu.gh
58
Cape Coast market at a value of GH₵8.50/Kg and GH₵11.87/kg respectively. The Cape
Coast and Half Assini had the highest variation of 17.35 and 16.83 respectively.
Variability in average Koobi Prices was quite low for between 2012 and 2017. It varied
between 8.16% for Tema to 17.35% for Cape Coast with the average variability being
computed at 14.09%. This implies that the price of Koobi in the markets analyzed did not
fluctuates widely from 2012- 2017. Table 4.1 presents the descriptive analysis of monthly
Koobi prices for the period 2012-2017.
Across the Kpala markets, the highest maximum price and highest average price of Kpala
was observed in the Cape Coast market at GH₵18.93/Kg and GH₵17.91/kg respectively.
The Lowest minimum price was observed in the Wa market at a value of GH₵ 8.33/kg
with the average price of GH₵ 11.59 /Kg observed in the Techiman market being the
lowest. Cape Coast and Half Assini had the highest variation of 17.35 and 16.83
respectively. Variability in Average Kpala prices was also quite low for the period (2012-
2017) under study. It varied between 5.35% for Makola to 9.60% for Ada with the
average variability being computed at 7.26%. Table 4.1 presents the descriptive analysis
of Kpala fish prices for the period 2012-2017.
In the smoked herring market, the Makola market had the highest maximum price and the
highest average price observed at GH₵19.21/Kg and GH₵15.08/kg respectively while the
minimum price was observed in the Kpandu market at GH₵ 7.62/kg and minimum
average whole sale price of GH₵ 11.77/kg was observed in the Cape Coast market. For
the smoked herrings, the Half-Assini and Kpandu market experienced the highest
variability in price at 16.91% while the least was 11.52% observed in the Tema market.
University of Ghana http://ugspace.ug.edu.gh
59
The average variability in smoked herring prices was recorded at 14.80%. Table 4.1
presents the descriptive analysis of smoked herring prices for the period 2012-2017.
The low variability in all the markets for processed fish implies that the price of Smoked
Herrings, Kpala and Koobi in the markets analyzed did not fluctuates widely from 2012-
2017. This means that changes in Smoked Herrings, Kpala and Koobi prices have been
quite stable for the time periods under study and in the markets analyzed. This can be
attributed to the ability of the processed fish to have a long shelf life than it fresh
counterparts making it available all year round. A high variability index would mean that
the prices of Koobi, Kpala and Smoked herring in the markets looked at fluctuated widely
between 2012 and 2017. High variability in prices means that producer incomes would be
unstable which is capable of causing unfavourable effects on production and production
planning (Mafimisebi, 2001). The effect of poor planning in production is can be harmful
to the welfare of consumers especially in countries like Ghana where poverty is still a
problem and expenditure on food makeup a larger portion of household’s disposable
income. This low variability in the prices of processed fish benefits both producers and
consumers of Smoked Herrings, Kpala and Koobi in the long run. Producers receive
stable incomes which positively influences the production and production planning
process which in turn improve the welfare of the consumer. These variations in prices can
be attributed to price fluctuations in the various markets under study which normally give
rise to temporal deficit from time to time.
University of Ghana http://ugspace.ug.edu.gh
60
Table 4.1 Descriptive statistics of monthly processed fish prices (2012-2017)
Koobi Kpala Smoked Herring
Markets Max Min Mean SD CV Max Min Mean SD CV Max Min Mean SD CV
Ada 17.58 10.50 13.76 2.25 16.35 13.04 8.91 12.08 1.16 9.60 17.32 9.61 13.98 2.16 15.45
Cape Coast 15.42 8.50 11.87 2.06 17.35 18.93 16.25 17.91 1.00 5.58 15.30 9.72 11.77 1.87 15.89
Half Assini 18.00 10.50 13.54 2.28 16.83 13.56 10.31 11.76 1.04 8.84 17.60 10.84 13.25 2.24 2.24 16.91
Makola 19.60 11.80 15.28 2.40 15.70 13.65 10.64 12.52 0.67 5.35 19.21 11.87 15.08 2.23 14.78
Mankessim 18.80 11.80 14.93 2.26 15.11 13.65 10.64 12.60 0.65 5.76 18.73 11.87 14.86 2.20 14.80
Techiman 17.80 11.00 14.33 1.74 12.14 13.06 10.13 11.59 0.79 6.81 15.17 10.28 13.37 1.54 11.52
Tema 15.20 11.00 13.74 1.10 8.16 13.18 9.57 12.06 0.83 6.88 15.69 10.31 13.17 1.60 12.15
Wa 17.00 11.10 13.49 1.71 12.67 12.91 8.33 11.67 1.08 9.25 17.35 10.59 13.42 1.76 13.11
Agbozume 17.80 10.80 13.70 2.25 16.42 - - - - - 18.75 10.53 14.00 2.31 16.50
Kpandu 22.20 10.80 12.73 1.30 10.21 - - - - - 16.25 7.62 13.25 2.24 16.91
Source: Author’s own computation from price data
University of Ghana http://ugspace.ug.edu.gh
61
4.1.2 Monthly Trend among markets for Processed Fish
Based on information on the prices of a product in the past, produces and consumers are able to
have an idea of what the price of a particular product will be in the future. Figure 4.1 shows the
visual plot of the monthly average of prices of Koobi from January 2012 to December 2017
across all regional markets considered.
The trend plots of Smoked herring, Koobi and Kpala show that these prices generally follow the
same path. They show a steady increase in the prices of processed fish over time. The bumper
harvest in fishing begins in July and ends around October sometime extending to November.
Prices in the market for Koobi, Kpala and smoked herring are characterized by periods of sharp
increases followed by periods where the prices remain constant and then jump again. This is
surprising considering we expect the price of processed fish to be low in the bumper season this
however was seen to occur mostly just in the Makola market for Processed fish. The Makola
market in Accra had the highest prices for both smoked herrings and Koobi. This may be due to
population and Accra serving as the capital of country and also may be due to Accra being an
area of high production and consumption of fish. In the case of the Kpala market, the Cape Coast
market experienced the highest prices.
The prices in the Cape Coast market were extremely high when compared to the prices of Kpala
prevailing in the other markets. This may be due to the Cape Coast market for Kpala not being
well-integrated with the other markets for Kpala. Thus the Cape Coast it is inefficient and
consumers for Kpala in Cape Coast are at a disadvantage compared to consumers in the other
markets. The jump in prices of processed fish can be attributed to the actions of processed fish
traders in general who intentionally hold back some of the processed fish during bumper periods
University of Ghana http://ugspace.ug.edu.gh
62
in order not to flood the market with too much fish and also have some to sell during the minor
fish season when prices are high thereby effectively controlling the prices of processed fish.
Figure 4.1, 4.2 and 4.3 show the trend plots for the monthly prices of Koobi, Kpala and smoked
herrings for the period under study
Figure 4.1: Trend plot of monthly prices of Koobi.
Figure 4.2: Trend plot of monthly prices of Kpala.
7.0
9.0
11.0
13.0
15.0
17.0
19.0
21.0
Jan
-12
Ap
r-1
2
Jul-
12
Oct
-12
Jan
-13
Ap
r-1
3
Jul-
13
Oct
-13
Jan
-14
Ap
r-1
4
Jul-
14
Oct
-14
Jan
-15
Ap
r-1
5
Jul-
15
Oct
-15
Jan
-16
Ap
r-1
6
Jul-
16
Oct
-16
Jan
-17
Ap
r-1
7
Jul-
17
Oct
-17
CapeCoast
Ada
Agbozume
HalfAssini
KpandorTorkor
Mankessim
Techiman
Tema
Wa
Makola
6.0
8.0
10.0
12.0
14.0
16.0
18.0
20.0
Jan
-12
Ap
r-1
2
Jul-
12
Oct
-12
Jan
-13
Ap
r-1
3
Jul-
13
Oct
-13
Jan
-14
Ap
r-1
4
Jul-
14
Oct
-14
Jan
-15
Ap
r-1
5
Jul-
15
Oct
-15
Jan
-16
Ap
r-1
6
Jul-
16
Oct
-16
Jan
-17
Ap
r-1
7
Jul-
17
Oct
-17
CapeCoast
Ada
Half Assini
Mankessim
Techiman
Tema
Wa
Makola
University of Ghana http://ugspace.ug.edu.gh
63
Figure 4.3: Trend plot of monthly prices of smoked herrings.
Source: Own computation from price data from Ghanaian Statistical Service.
4.2 Unit Root Test Results
All the price data for Kpala, Koobi and Smoked herring from 2012 to 2017 were tested for
stationarity as a prerequisite for Cointegration analysis. The Augmented Dickey Fuller (ADF) test
was used to test the null hypothesis of non-stationary for all the price series. Table 4.2 presents
the results of the stationarity test for the markets in the sampled period. The appropriate lag
length was selected based on the comparison of the Hannah-Quinn criterion, Akaike Information
Criteria and Schwarz Bayesian Criterion.
4.2.1 Unit Roots Test Results for Processed Fish
The results of the ADF test show that at the 1% critical values of -3.552, the null hypothesis of
unit root in the Koobi market cannot be rejected for all price series except in the case of Kpandu.
8.0
10.0
12.0
14.0
16.0
18.0
20.0
22.0
24.0Ja
n-1
2
Ap
r-1
2
Jul-
12
Oct
-12
Jan
-13
Ap
r-1
3
Jul-
13
Oct
-13
Jan
-14
Ap
r-1
4
Jul-
14
Oct
-14
Jan
-15
Ap
r-1
5
Jul-
15
Oct
-15
Jan
-16
Ap
r-1
6
Jul-
16
Oct
-16
Jan
-17
Ap
r-1
7
Jul-
17
Oct
-17
CapeCoast
Ada
Agbozume
HalfAssini
KpandorTorkor
Mankessim
Techiman
Tema
Wa
Makola
University of Ghana http://ugspace.ug.edu.gh
64
The null hypothesis of non-stationarity is rejected after testing the first difference of all price
series for the presence of unit root. The null hypothesis of no unit roots (that is the series is
stationary) in the level of the price series at the 1% and 5% significance levels is strongly rejected
for all markets except Kpandu which was already Stationary at level, but cannot reject the null
hypothesis at the first difference of the price series. The results therefore show that all the price
series of Koobi were created by the same stochastic processes and therefore display the
propensity toward achieving equilibrium in the long run except the Kpandu price series (Chirwa,
2001; Mafimisebi, 2001). Cointegration can however be tested for between Makola and the other
markets except in the case of the Kpandu market. The different order of integration between the
Makola market and the Kpandu markets lead to the conclusion of an absence of integration
between them (Rapsomanikis et al., 2006).
The Null hypothesis of non-stationarity at the conventional level of 5% could not be rejected for
all price series in the Kpala market at their level except for the price series of the Ada, Tema and
Wa markets. This implies that the prices of Tema, Ada and Wa were found to be stationary at
their levels. After taking a first difference and testing for stationarity the null hypothesis is
rejected except in the case of the Tema, Ada and Wa markets. The prices series in the Kpala
market were found to be integrated of order one I (1) except for Ada, Tema and Wa which were
integrated of an order zero. The Kpala markets prices series were integrated of different orders. .
in the case of the Kpala market too, the price series created by similar stochastic processes also
display the propensity to achieve a long-run equilibrium (Mafimisebi, 2012; Nielsen, 2006).As
such cointegration analysis between Makola and the other markets can be carried out for all price
series except the Ada, Tema and Wa. We conclude the absence of integration between the
University of Ghana http://ugspace.ug.edu.gh
65
Makola market and the Ada, Tema and Wa markets as they were integrated of different orders
(Rapsomanikis et al., 2006).
In the case of the smoked herring market, the results of the ADF test reveals that the null
hypothesis of unit root, H0:ρ = 0 that is the price series is non- stationary, cannot be rejected for
all price series at 1%, 5% and 10% critical values. The null hypothesis is however rejected after
taking a first difference of all price series and testing for the presence of unit roots. The price
series under study are thus a first-difference stationary process which implies that they have unit
root or are integrated of order one I(1) (Gujarati, 2008). The result shows that the price series of
smoked herrings were created by the same stochastic processes and therefore show the tendency
to move towards achieving equilibrium in the long run. This allow for co-integration tests for the
testing of the long run equilibrium relationship to be carried out.
This result of the unit root for Koobi, Kpala and smoked herring is backed by previous findings
that price of most food commodity are non-stationary in nature at their levels and stationary after
first-differencing (Okoh & Egbon, 2003; Chirwa 2001; Mafimisebi, 2001 and Oladapo, 2003).
This could be due to the trends in the series caused by inflation and cyclical variations from
season leading to mean non-stationarity.
University of Ghana http://ugspace.ug.edu.gh
66
Table 4.2: Results of ADF tests on the monthly processed fish price series
ADF (Koobi) ADF (Kpala) ADF (Smoked Herring)
Markets Levels First Difference
Levels First Difference Levels First Difference
Ada 0.183 -6.746*** -3.871*** -5.928 -1.746 -6.488***
Cape Coast -0.561 -6.510*** -1.029 -6.148*** 1.056 -5.614***
Half Assini 0.404 -6.122** -0.445 -6.182*** 0.747 -6.745***
Mankessim 0.029 -6.571*** -2.477 -6.612*** 0.043 -6.358***
Techiman 0.772 -7.346*** -0.747 -7.705*** -1.529 -7.202***
Tema -2.576 -6.397*** -3.943*** -9.716 -0.954 -6.920***
Wa 0.890 -6.983*** -5.363*** -7.541 0.346 -6.518***
Makola -0.137 -5.370*** -2.751 -6.932*** -0.196 -5.733***
Agbozume 0.443 -7.002*** - - 0.388 -7.579***
Kpandu -4.577*** -9.716 - - -1.909 -9.576***
Source: Author’s own computation from price data
*** means rejection of the null hypothesis at the 1% significant level
University of Ghana http://ugspace.ug.edu.gh
67
4.3 Cointegration Test Results
With the confirmation of the price series being stationary in their first difference, the existence of
co integration among the selected markets pairs was tested using the Johansen’s multiple co
integration method. As indicated in chapter 3 Co-integration between two non-stationary price
series implies that a linear combination of the two series is stationary and the prices, therefore,
tend to move together or follow the same path in the long-run. Since the results from the ADF
test revealed different orders of integration for some of the price series in the Koobi and Kpala
market when compared to the Makola market. Cointegration analysis could not be carried out for
all market pairs.
4.3.1 Extent of Market Integration between Processed Fish Market
Cointegration was used to determine the extent of market integration for the processed fish under
study. The result for the test of cointegration among the Smoked Herring, Koobi and Kpala
markets reveal that out of the nine market pair subjected to a cointegration test only two were
found to be cointegrated. Evidence of cointegration was found to exist between the Makola-
Mankessim and Makola-Kpandu market pairs. This implied that 22.22% of smoked herring
market pairs were co-integrated of the order one I (1) at the 1% significance level. This shows the
number of smoke herring market pairs that have prices which follow the same path in the long
run despite divergence in the short-run.
In the markets for Koobi, out of the eight market pairs subjected to a cointegration test, only one
market pair rejected the null hypothesis of no co-integration at the 1% significant level.
Cointegration was found to exist only in the Makola-Mankessim market pair at the 1%
significance level. This means that only 11% of Koobi markets were found to have prices which
co-move in the long run despite divergences that occur in the short-run due to price shocks. Eight
University of Ghana http://ugspace.ug.edu.gh
68
instead of nine market pairs were tested for cointegration due to the conclusion of an absence of
integration between the Makola-Kpandu pair. This was done following the framework of
Rapsomanikis et al. (2006). The results of the Johansen cointegration test for smoked Koobi are
presented in table 4.3.
Out of a possible a seven market pairs that could have been tested for cointegration, only four
market pairs were subjected to the test. This was because of the different orders of integration
between the Makola-Ada, Makola-Tema and Makola-Wa hence an absence of integration
(Rapsomanikis et al., 2006). There was no evidence of cointegration among the market pairs for
Kpala. This implies that 0% of the markets for Kpala had their prices co-moving together in the
long run. The results of the Johansen cointegration test for Kpala are presented in table 4.3.
Thus, there is a long run relationship among the Makola-Mankessim and Makola-Kpandu
markets pairs for smoked herring and among the Makola-Mankessim market pair for Koobi. In
other words, in the event of a price shock, prices in the cointegrated market may differ from each
other in the short run but will not drift apart in the long run. This suggests that those market pairs
may be efficient since market integration has been used as a measure of market efficiency. The
findings imply that similar stochastic processes, possibly induced by efficient information flow,
drive the dynamics of prices in the system of markets (Motamed et al., 2008). The results also
reveal that most of the markets for processed fish are not integrated or interact with each other.
The co-movement of price in markets over period of time can serve as indications of well
integrated markets, which imply that the markets are functioning proper and are efficient. The
nonexistence of integration in markets may pass on incorrect price signals that might alter the
marketing decisions taken by producers resulting in inefficient product movement (Goodwin &
Schroeder, 1991). In a common domestic processed fish market in Ghana, prices between
University of Ghana http://ugspace.ug.edu.gh
69
markets adjust towards achieving long-run equilibrium. This however, was not the case for most
of the processed fish markets analyzed. Causality of the price series was evaluated with the
Granger causality test at 1 lag for all the markets for processed fish after the test of cointegration.
The results for the test for cointegration among the processed fish markets pairs is presented in
table 4.3.
Table 4.3 Johansen cointegration test Statistics for processed fish markets
Source: Own computation from processed fish price data for 2012-2017.
The asterisks *** denote rejection of the null hypothesis of no cointegration vector at the1%. The
critical values for r = 0 and r = 1 at the 1% significance levels are 20.04 and 6.65
Market Pairs Null
Hypothesis
Trace Statistics
(Smoked herring)
Trace
Statistics(Koobi)
Trace Statistics
(Kpala)
Makola-Cape Coast r = 0
r ≤ 1
18.09
0.48
6.99
0.02
12.53
4.424
Makola-Half Assini r = 0
r ≤ 1
16.45
0.39
8.89
0.23
17.20
0.95
Makola-Mankessim r = 0
r ≤ 1
24.01***
0.00
22.25***
0.03
17.35
4.94
Makola-Techiman r = 0
r ≤ 1
8.18
1.40
15.68
0.06
11.94
1.39
Makola-Ada r = 0
r ≤ 1
13.94
0.51
10.19
0.07
Absence of
integration
Makola-Tema r = 0
r ≤ 1
14.33
0.06
11.95
0.64
Absence of
integration
Makola-Wa r = 0
r ≤ 1
11.63
0.21
13.64
0.00
Absence of
integration
Makola-Agbozume r = 0
r ≤ 1
12.93
0.15
15.64
0.16
-
Makola-Kpandu r = 0
r ≤ 1
26.18***
0.12
Absence of
integration
-
University of Ghana http://ugspace.ug.edu.gh
70
4.4 Direction of causality in Processed Fish Market
Granger causality is important in showing the direction of the relationship after performing co-
integration tests. This section shows the various casual relationships for all the price series. Table
4.4 shows the causal relationship between processed fish markets from 2012 to 2017 by applying
the granger causality test.
4.4.1 Result of Granger-causality Test for Processed Fish Markets
The granger causality test for Koobi revealed that five markets exhibited unidirectional causality,
3 markets exhibited bidirectional causality and 1market exhibited independent causality. For the
Smoked herring market, five market pairs exhibited unidirectional causality and 4 exhibited
bidirectional causality. There was no independent causality in the smoked herring markets. In the
markets for Kpala three markets pairs exhibited unidirectional causality, 2 exhibited bidirectional
causality and two exhibited independent causality.
The presence of bidirectional causality in some of the markets for processed fish implies that
those markets are strongly integrated with each other, experiencing physical arbitrage to settle
any disequilibrium between the markets. In the market links that showed a unidirectional (one
way) granger-causality that is there is no significant causality from the other market (Hendry,
1986). As reported in Fackler (1996), Gupta & Mueller (1982) argue that the failure of one price
to be predictive of another when the second is predictive of the first (unidirectional causality) is
an indication that the second price is not incorporating the price information from the first region.
Unidirectional causality is therefore taken to indicate that a market is inefficient in terms of
information. The independent causality implies that the market pairs are therefore independent
and autonomous do not incorporate the price information from one another in setting their prices
suggesting that these markets may be inefficient. The presence of independent causality among
University of Ghana http://ugspace.ug.edu.gh
71
the market pairs does not imply a total absence of price transmission in the market links. This
might just mean that price signals are transmitted instantaneously under special conditions like
storage, inventory holding and delays in transportation. Results of the granger-causality are
presented in table 4.4 below.
Table 4.4: Result of granger-causality test for processed fish markets
Source: Own computation from Koobi price data for 2012-2017.
***,**,* represents rejection of null hypothesis at 1%.5% and 10% significance level
Koobi Smoked Herring Kpala
Null
Hypothesis
F-Stats
Comment F-Stats
Comment F-Stats Comment
MK-AD
AD-MK
2.15
8.08***
Unidirectional 6.39**
2.26
Unidirectional 3.93**
0.47
Unidirectional
MK-CC
CC-MK
3.50**
3.49**
Bidirectional 1.28
6.33**
Unidirectional 5.77**
0.06
Unidirectional
MK-HA
HA-MK
1.16
8.19***
Unidirectional 4.56**
8.44***
Bidirectional 3.92**
5.60**
Bidirectional
MK-MN
MN-MK
2.37
7.42***
Unidirectional 4.67**
5.67**
Bidirectional 1.72
0.14
Independent
MK-TE
TE-MK
9.62***
1.61
Unidirectional 1.09
3.40*
Unidirectional 1.53
2.70
Independent
MK-TM
TM-MK
6.30**
1.66
Unidirectional 3.50*
3.26*
Bidirectional 3.10*
7.78***
Bidirectional
MK-Wa
Wa-MK
8.47***
4.26**
Bidirectional 10.47***
2.68
Unidirectional 2.50
2.80*
Unidirectional
MK-AG
AG-MK
4.08**
8.59***
Bidirectional 5.64**
7.44***
Bidirectional
MK-KP
KP-MK
0.28
1.28
Independent 0.07
6.13**
Unidirectional
University of Ghana http://ugspace.ug.edu.gh
72
4.5 Price Transmission between Processed Fish Markets in Ghana.
The evidence of significant cointegrating vectors between the market pairs for processed fish is a
necessary condition for using the VECM to determine the effects of price shocks on price
adjustment. However since the VECM is embedded in the AVECM, the AVECM will be used to
evaluate the impact of price shocks on price adjustment as well as the presence and extent of
asymmetry. The Absence of such significant cointegrating vectors requires the specification of a
ARDL model to analyze price adjustment. The results of the econometric estimation of the
VECM for the producer and consumer market pairs are presented in Table 4.5and that of the
ARDL in table 4.6.
4.5.1 Results of the Asymmetric Vector Error Correction Model
It is believed that markets are characterized by asymmetry in the price transmission process
where traders react faster to shocks that reduce their marketing margin than to those that stretch
them. The evidence of asymmetry in transmission of price signals between the central market and
consumer markets for processed fish was estimated using the Asymmetry Vector Error
Correction Model and the results are presented in Table 4.5.
In the Smoked Herring Market, the Makola-Mankessim market pair, following a positive shock
that creates disequilibrium, 50 % of such shocks will be eliminated within a month and it will
take only 1.01months for the system to return to equilibrium. In the events of a negative shock
46% of such deviations will be corrected within a month and the system returns to equilibrium in
1.13 months. The results also reveal that the is asymmetry occurring in the price transmission
between the Makola- Mankessim market pair implying that positive shocks were transmitted
faster than negative shocks and not vice versa. This observation from the analysis confirmed the
assertion of farmers that producer price increases are passed on faster to consumers than producer
University of Ghana http://ugspace.ug.edu.gh
73
price decreases. The causes of asymmetry in price transmission between some of the markets may
depend on the characteristics of the smoked herring market. However, considering the abilities of
traders and their associations to influence the conduct of the market by determining how much to
release into the market (Langyintuo, 2010), inventory management and stock behaviour
potentially stands as a motivating cause of asymmetry.
Abdulai (2000) rules out menu cost as a cause of asymmetry in the smoked herring market since
prices in the markets are determined through personal negotiation between producers/traders and
consumers. This is the case in most markets for agricultural produce. In the case of menu and
search cost as a cause of asymmetry the rapid increase in the use of mobile phones and computers
makes it easier for information to accessed online and transmitted at a faster rate than was the
case in the past and as menu and search cost incurred by consumer are drastically reduced making
information asymmetry a minimal option in causing asymmetry in price transmission in Ghana.
Also, the government has not been actively or directly involved in trading and pricing of smoked
herring. However the Government of Ghana is actively seeking international cooperation to assist
the country in further aquaculture development in the country. Efforts are being made to
modernize the fisheries sector and this might have an effect on price in the coming years as such
government intervention is not a suitable candidate for causing asymmetry in price transmission.
In The Makola-Kpandu market pair, 52% of the disequilibrium caused by positive shocks is
corrected in a month and the system returns to equilibrium in 0.95 months. In the case of negative
shocks to the system 67% of the created disequilibrium is corrected in month and the system
returns to equilibrium in 0.41 months. The absence of asymmetry in the markets can be taken to
University of Ghana http://ugspace.ug.edu.gh
74
mean the nonexistence of imperfections present in the market Ben-Kaabia et al. (2002). Thus the
Makola-Kpandu market for smoked herrings is efficient.
The speeds of adjustment for positive and negative deviations in the Makola-Mankessim and
Makola-Kpandu for smoked herring are higher than that of Mensah-Bonsu et al. (2011). Mensah-
Bonsu et al. (2011) found a speed of adjustment of 27.73% and summed up that there is weak
integration among the markets. Thus it can be concluded that there is a relatively strong
integration among the markets. As such cointegrated smoked herring markets may be able to
respond more quickly to market shocks, and market channel members can efficiently and
effectively distribute processed fish from surplus to deficit markets.
In the Koobi Market, the Makola-Mankessim market pair, following a positive shock that creates
disequilibrium, 39 % of such shocks will be eliminated within a month and it will take only 1.42
months for the system to return to equilibrium. In the events of a negative shock 24% of such
deviations will be corrected within a month and the system returns to equilibrium in 2.92 months.
The results also reveal that the is symmetry occurring in the price transmission between the
Makola- Mankessim market pair implying that positive and negative shocks to price were
transmitted at the same speed and time. The existence of asymmetry in the transmission of price
shocks means that there is some loss in welfare for a number of market participants since the
distribution of welfare could be different under symmetry (Wlazlowski et al., 2009). This is
however not the case in the Makola-Mankessim market pair for Kpala which was characterized
by symmetric relationship.
From Table 4.5 the speed of adjustment of 39% for positive shocks and 24% for negative
deviations in the Makola-Mankessim for Koobi indicate there is weak integration among the
University of Ghana http://ugspace.ug.edu.gh
75
markets. This is the same as the conclusion drawn by Mensah-Bonsu et al. (2011). Mensah-
Bonsu et al. (2011) found a speed of adjustment of 27.73% and summed up that there is weak
integration among the markets. Thus, an improvement in the accuracy of information to market
participants will improve market integration. This will cause Koobi markets to respond more
quickly to market shocks, and market channel members can efficiently and effectively distribute
Koobi from surplus to deficit markets improving upon food security.
Table 4.5 Results of the AECM Model
Smoked Herrings Koobi
Market
Pairs 𝐸𝐶𝑇𝑡−1
− 𝑇ℎ𝑎𝑙𝑓 𝐸𝐶𝑇𝑡−1+ 𝑇ℎ𝑎𝑙𝑓 𝐴𝑠𝑦
𝐸𝐶𝑇𝑡−1 − 𝑇ℎ𝑎𝑙𝑓 𝐸𝐶𝑇𝑡−1
+ 𝑇ℎ𝑎𝑙𝑓 𝐴𝑠𝑦
Makola-
Mankessim
-0.46** 1.13 -0.50*** 0.41 0.01 -0.24 2.92 -0.39** 1.42 0.17
Ma kola-
Kpandu
-0.67*** 1.01 -0.52** 0.95 0.22
Source: Own computation from processed fish price data (2012-2017).
***,**,* represents the significance of the coefficients at the 1%, 5% and 10 % significance
level
4.5.2 Evidence of Price Transmission (ARDL Model)
The results of the ADRL reveal that in the Koobi market only the coefficient of the previous price
of Makola, Current and previous price of Techiman and the current price of Agbozume were
significant. This implies that only these prices explain the current prices of Makola. A 1%
increase in the previous price of Makola will increase the current price of Makola by 0.62%. A
1% increase in the current price of Techiman will increase the current Makola price by 0.59%
whereas in the previous price of Techiman will decrease the Makola price by 0.37%. In the
Agbozume market a 1% increase in the current price will increase the current Makola price by
0.41%. The effect of the other prices on the current Makola price was insignificant. Thus at any
University of Ghana http://ugspace.ug.edu.gh
76
point in time, pricing of Koobi depends on the prevailing market condition and the previous price
of Makola, Current and previous price of Techiman and the current price of Agbozume since
those values gives an explanation of the current prices of Koobi in the Makola market.
In the Kpala market, only the previous price of Makola, current and previous price of Mankessim,
current price of Cape Coast and current price of Techiman explain the current price of Makola. A
1% increase in the previous price of Makola will increase the current Makola price by 0.50%. A
1% increase in the current Mankessim price increases Makola price by almost 1% whereas the
previous price of Mankessim will decrease the Makola price by 0.28%. A percent increase in the
current prices of Techiman will increase the price of Makola by 0.18%. An increase in current
Cape Coast prices for Kpala will reduce the current prices prevailing in Makola by 0.18%. As
such, at any point in time, pricing of Kpala depends on the prevailing market condition and only
the previous price of Makola, current and previous price of Mankessim, current price of Cape
Coast and current price of Techiman since those values explain the current prices of Kpala in the
Makola market. Acquah et al. (2012) found out that only the present prices of cassava in the
Kumasi market give an explanation of the present prices prevailing in the Techiman market at 5
percent level of significance. The past price values of all the markets were insignificant showing
that they were not useful in explaining the current prices of cassava in Techiman market
revealing.
In the market for smoked herring, the previous price of Makola, current and previous price of
Cape Coast, current price of Ada and current price of Tema influence the current price of Makola.
A 1% increase in the previous price of Makola will increase the current Makola price by 0.53%.
A 1% increase in the current price of Cape Coast will increase the current Makola price by 0.80%
whereas the previous price of Cape Coast will decrease the current Makola price by 0.52%. A
University of Ghana http://ugspace.ug.edu.gh
77
percent increase in the current price of Ada will decrease the price of Makola by 0.37%. An
increase in current Tema prices for Smoked Herrings will lead to a 0.30% rise in the current
prices prevailing in the Makola market. Thus at any point in time, pricing of Smoked herring will
depend on the conditions prevailing in the market and also on the previous price of Makola,
current and previous price of Cape Coast, current price of Ada and current price of Tema since
those values explain the current prices of Koobi in the Makola market. These results are similar to
a study by Acquah et al. (2012) on the market integration and price transmission of cassava
markets in Ghana.
Table 4.6 Results of the ARDL model
Markets Koobi Smoked Herring Kpala
Coef. Coef. Coef.
Makola L1. 0.62*** 0.53*** 0.50***
_cons 0.07 0.02 0.20
Mankessim - - 0.90***
Mankessim L1 - - -0.28**
Cape Coast -0.07 0.80*** -0.18*
Cape Coast L1. 0.08 -0.52** -0.07
Techiman 0.59** 0.05 0.18**
Techiman L1. -0.37* -0.01 -0.07
Half Assini 0.03 0.16 0.03
Half Assini L1. -0.03 -0.11 -0.01
Ada 0.08 0.20
Ada L1. 0.04 -0.37**
Agbozume 0.41* -0.03
Agbozume L1. -0.21 0.05
Tema -0.01 0.30*
Tema L1. -0.03 -0.05
Wa -0.11 -0.02
Wa L1. -0.03 0.02
Kpandu -0.03 -
Kpandu L1. 0.03 -
Source: Authors’ construct.
***,**,* represents the significance of the coefficients at the 1%, 5% and 10 % significance
level
University of Ghana http://ugspace.ug.edu.gh
78
CHAPTER FIVE:
SUMMARY, CONCLUSSIONS AND RECOMMENDATIONS
5.0 Introduction
The study is summarized in this chapter. Conclusions based on the research results are also
presented in this chapter. Policy recommendations based on the findings and conclusions of the
study are also presented in this chapter.
5.1 Summary and Major Findings
This study sought to examine market integration and price transmission between some processed
fish markets in Ghana. The main objective of this study was to examine the efficiency of
processed fish marketing system in Ghana. The specific objectives being to analyze the extent of
Market integration among the selected processed fish markets in Ghana, to analyze the extent of
price transmission among selected processed fish markets in Ghana and to analyze the extent of
price asymmetry in selected processed fish markets in Ghana using monthly processed fish prices
data between 2012 and 2017. The Stata 14 and JMulti softwares were used to analyze the data.
A descriptive analysis of the data shows that in both the market for Koobi and smoked herring
the highest average values for monthly prices was found in the Makola market while the lowest
value was observed in the Cape Coast Market. However in the Market for Kpala the highest
average values for monthly prices was found in the Cape Coast market while the lowest value
was observed in the Techiman Market. The variability in processed fish market prices as
estimated by the coefficient of variation was on the average was found to be 14.09% for Koobi,
14.80% for smoked herring, 7.26% for Kpala. The low fluctuations in mean prices indicate that
processed fish prices were relatively stable. A trend analysis of all the markets for processed fish
University of Ghana http://ugspace.ug.edu.gh
79
showed a steady and persistent increase in prices with very little variation over the period 2012 to
2017.
All the price series in the Koobi market where integrated of the order one I (1) meaning that
similar stochastic processes created these series except the price series for Kpandu which was
stationary at level. Those of Kpala were all integrated of order I one (1) except in the case of Wa,
Tema and Ada. The price series for smoked herrings were all integrated of the order one I (1) at
the 1% level also indicating they were created by similar stochastic processes. Evidence of
cointegration was found to exist in the Makola-Mankessim market pair for Smoked herring and
Koobi and Makola-Kpandu market for smoked herring. There was no cointegration found in any
of the market pairs for Kpala. The presence of bidirectional causality indicates that some of the
markets are well integrated with the Makola market. Unidirectional causality in the markets links
for processed fish show the markets are not strongly integrated the presence of independent
causality among some market pairs in the Koobi and Kpala markets indicate some inefficiency in
those market pairs.
The results of the AVECM indicate that Positive shocks were corrected faster than negative
shocks in the Makola-Mankessim market pair for smoked herring and Koobi and in the Makola-
Kpandu market for smoked Herring. The speed of adjustment in the smoked herring markets
when compared to a perfect adjustment of 100% threshold was relatively Strong. The case of the
Koobi market was relatively weak. Also the adjustment mechanism for these markets after a
shock was characterized by asymmetry in the Makola-Mankessim market pair for smoked
herring. The other market pairs had adjustment mechanisms that were characterized by symmetric
relationships. The results of the ARDL indicate that in the short run, the lagged values of the
Makola price had a positive and highly significant effect on the current price of processed fish in
University of Ghana http://ugspace.ug.edu.gh
80
the Makola market. In the Koobi market only the current price of Techiman and Agbozume and
the previous price of Techiman had a significant effect on the current Makola price previous price
of Ada and current price of Tema. The current prices of Kpala in Mankessim, Cape Coast and
Techiman and the previous price of Mankessim had a significant effect on the current price of
Kpala in the Makola market. For the smoked herring markets, the current price of Makola was
influenced by the current price of Cape coast and Tema, and the previous price of Cape Coast and
Ada.
5.2 Conclusions
Low variability in prices imply that price of processed fish did not fluctuate widely across
seasons in all markets analyzed. This translates into stable producer incomes mostly for the
women who are engaged n the processing of fish.
The results revealed all the price series of smoked herring, nine out of ten for koobi and five out
of eight for Kpala markets in Ghana are integrated of order one. The results of the Johansen
cointegration analysis show that most of the market pairs for all the processed fish studied are not
cointegrated. The absence of cointegration therefore indicates that transmission of price signals is
poor among the processed fish markets in Ghana, leading to market inefficiencies that need to be
addressed. Additionally, there is a mixture of bidirectional, unidirectional and independent
causality between Makola the other markets implying that some of the markets are not very well
integrated. Overall the results of the extent of market integration analysis obtained by the
Johansen Maximum Likelihood approach indicate that the markets for smoked herring, Kpala and
koobi were not well integrated.
University of Ghana http://ugspace.ug.edu.gh
81
From the results of the AECM, there is relatively weak integration/adjustment in the Koobi
market as compared to a relatively strong adjustment for smoked herring. As such in the Koobi
market there is the need for further market integration to improve the price transmission process.
Also asymmetry was found only in the Makola-Mankessim market for smoked herrings. The
results of ARDL also show that most of the current and previous prices of the markets did not
influence the current prices in the Makola market. Thus in most cases, the price of the processed
fish depends on the condition prevailing in the market at that point in time. The previous price of
the Makola market was however significant in explaining the current price of Makola for Koobi,
Kpala and Smoked Herring.
In markets that were cointegrated, the adjustment mechanism after a shock was characterized
mostly by symmetry. There was therefore no asymmetry in the markets, thus these particular
markets are efficient and well-integrated. Overall it can be concluded from the results that the
markets for processed fish are not very efficient. There is therefore the need to improve the
efficiency of these markets through improving market integration and price transmission of
processed fish markets. This will ensure fish moves from surplus areas to deficit areas thus
improving food security.
5.3 Policy Recommendations
Most of the markets for processed fish in Ghana were found to be mostly inefficient in
performing their functions. Thus it is recommended that the Ministry of Road transport and the
Ministry of Communication together can improve marketing infrastructures such as roads and
communication facilities. This will greatly reduce transaction costs and improve price
transmission and market integration in the processed fish market thus increasing market
efficiency.
University of Ghana http://ugspace.ug.edu.gh
82
Since the adjustment mechanism after a shock for the Makola-Mankessim market pair for
Smoked Herring was characterised by asymmetry which are signals market failure, redistribution
and net welfare losses to producers and consumers. It is recommended that the potential causes of
asymmetry such as the ability of fish mummies to influence prices and the inventory and stock
behaviour of traders be looked into. This can be done through investing in storage facilities such
Cold storages or processing plants by the Government and the Ministry of Fishery and
Aquaculture Development given the seasonal nature of the commodity.
Also since the Ghana Statistical Service collects market information on the prices of processed
fish in the various markets, the information should be timely and made available to the producers
of processed fish. This will allow them to know which markets offer remunerative prices for their
products and lead to efficient arbitrage between markets. This will lead to market systems that are
efficient in performing their functions; where most markets are integrated with each other
ensuring processed fish is delivered to consumers at competitive cost.
University of Ghana http://ugspace.ug.edu.gh
83
REFERENCES
Abdulai A. 2002. Using threshold co integration to estimate asymmetric price transmission in the
Swiss pork market. Journal of Applied Economics.34: 679– 687.
Abdulai, A. (2000). Spatial price transmission and asymmetry in the Ghanaian maize market
Journal of Development Economics, 63(2), 327-349.
Abdulai, A. (2007). ‘Spatial and vertical price transmission in food staples market chains in
eastern and southern Africa: What is the evidence?’ Paper presented at the FAO trade and
markets division workshop on staple food trade and market policy options for promoting
development in eastern and southern Africa. Rome, March 1-2.
Abunyuwah, I. (2007). ‘Market integration analysis and time series econometrics-conceptual
insights from Markov-switching models’. (Doctoral dissertation, Faculty of Agricultural
Sciences, Georg-August-university of Göttingen, Germany, 2007).
Aburajab, T. (1999). Investigating production opportunities, marketing efficiency, and options of
trade for fruits and vegetables in Palestine. Verlag Ulrich E. Grauer.
Acquah, De-G. H., Micah, J.A., & Owusu, R., 2012. Price Transmission in Domestic Agricultural
Markets, An Evidence from Selected Cassava Markets in Ghana. Journal of Sustainable
Development in Africa. 14(5)
Acquah, De-G.H & Owusu, R. 2012.Spatial Market Integration and Price Transmission of
Selected Plantain Markets in Ghana. Journal of Sustainable Development in Africa.14(5).
Adenegan, K. O., & Bolarinwa, A. O. (2010).Price Transmission and Market Integration of Fish
in Oyo State. Journal of Rural Economics and Development, 19(1).
University of Ghana http://ugspace.ug.edu.gh
84
Adetunji, M. O., & Adesiyan, I. O. (2008). Economic analysis of plantain marketing in Akinyele
local government area in Oyo state, Nigeria. International Journal of Agricultural
Economics and Rural Development, 1(1), 15-21.
Aggrey-Fynn, E.2001.The contribution of the fisheries sector to Ghana's economy. A paper
prepared on behalf of the FAO as an input into the Sustainable Fisheries Livelihoods
Study. April 2001.
Alderman, H. & Shively, G. E. (1991). ‘Prices and markets in Ghana’.Cornell Food and Nutrition
Policy Program Working Paper, No. 10, Ithaca, NY.
Alexander, C. (2001). Market models: A guide to financial data analysis. John Wiley & Sons.
Alexander, C., & Wyeth, J. (1994). Cointegration and market integration: An application to the
Indonesian rice market. The Journal of Development Studies, 30(2), 303-334.
Amikuzuno, J. (2010). Spatial price transmission and market integration between fresh tomato
markets in Ghana: Any benefits from trade liberalization? Paper Presented To Department
of Agricultural Economics and Extension, University for Development Studies, Tamale,
Ghana.
Amonde, T., McDonald, L. & Barrett, K. (2009).An economic enquiry into the causes of the
perceived asymmetric price transmission in markets for specific consumer goods in
Jamaica. Paper presented to the fair trading commission (FTC) of Jamaica, April, 2009.
Ankamah-Yeboah, I. (2012). Spatial price transmission in the regional maize markets in Ghana.
Anon. (2008). “The fishing sub- sector and Ghanaian Economy” The Bank of Ghana sector
studies on fisheries, Research Department, 2008. 5-11pp
University of Ghana http://ugspace.ug.edu.gh
85
Ardeni, P. G. (1989). Does the law of one price really hold for commodity prices?. American
Journal of Agricultural Economics, 71(3), 661-669.
Awity, L. 2005. Prospective analysis of future aquaculture development, National aquaculture
sector review, Ghana
Bada, T., & Rahji, M. A. Y. (2010). Market delineation study of the fish market in Nigeria: An
Application of Cointegration Analysis. Journal of Agricultural Science, 2(3), 158.
Baffes, J. (1991). Some further evidence on the law of one price: The law of one price still
holds. American Journal of Agricultural Economics, 73(4), 1264-1273.
Bailey, D., & Brorsen, B. W. (1989). Price asymmetry in spatial fed cattle markets. Western
Journal of Agricultural Economics, 246-252.
Balcombe, K.G. & Morrison, J. 2002. Commodity price transmission: A critical review of
techniques and an application to selected export commodities. Report to the Food and
Agriculture Organization of the United Nations.
Balke, N. S., Brown, S. P., & Yucel, M. K. (1998). Crude oil and gasoline prices: an asymmetric
relationship?. Economic Review-Federal Reserve Bank of Dallas, 2.
Ball, L. & Mankiw, N. G. (1994). Asymmetric price adjustment and economic fluctuations. The
Economic Journal, 104 (423), 247–261.
Bank of Ghana. (2008). The fishing sub sector and Ghana’s Economy. (Online), Available
http://www.bog.gov.gh/privatecontent/Research/Sector%20Studies/fisheries_completerp
df.pdf. Accessed on 2nd june, 2018.
University of Ghana http://ugspace.ug.edu.gh
86
Bannor, R. K., & Bentil, J. K., 2014.Cassava Markets Integration Analysis in the Central Region
of Ghana.Indian Journal of Economic Development. 10(4): 319-329
Barrett, C. (1996). Market analysis methods: are our enriched toolkits well suited to enlivened
markets? American Journal of Agricultural Economics, 78, 825–9.
Barrett, C. 2001. Measuring integration and efficiency in international agricultural markets.
Review of Agricultural Economics 23: 19–32.
Barrett, C. B. & Li, J. R. (2002).Distinguishing between equilibrium and integration in spatial
price analysis.American Journal of Agricultural Economics, 84, 292-307.
Barrett, C. B. (2005). Spatial market integration. The new Palgrave dictionary of economics (2nd
ed.), London, Palgrave Macmillan.
Baulch, B. (1997). Transfer costs, spatial arbitrage, and testing for food market
integration. American Journal of Agricultural Economics, 79(2), 477-487.
Béné, C., Barange, M., Subasinghe, R., Pinstrup-Andersen, P., Merino, G., Hemre, G. I., &
Williams, M. (2015). Feeding 9 billion by 2050–Putting fish back on the menu. Food
Security, 7(2), 261-274.
Bender, D.A. (1982). Dictionary of Nutrition and Food Technology. 5th edition. Cambridge, UK.
CRC Press.
Ben-Kaabia, M., Gil, J. M. & Boshnjaku, L. (2002).Price transmission asymmetries in the
Spanish lamb sector.International congress, August, 28-31, Zaragoza, European
Association of Agricultural Economists.
University of Ghana http://ugspace.ug.edu.gh
87
Borenstein, S., A. C. Cameron, & R. Gilbert. (1997) “Do Gasoline Prices Respond
Asymmetrically to Crude Oil Price Changes?” Quarterly Journal of Economics
112(1997):305–339. Doi: 10.1162/003355397555118.
Bukenya, J. O., & Ssebisubi, M. (2014). Price integration in the farmed and wild fish markets in
Uganda. Fisheries science, 80(6), 1347-1358.
Chirwa, E. W. (2000). “Liberalization of Food Marketing and Market Integration in
Malawi”.University of Malawi. Final Report Presented at the African Economic Research
Consortium Biannual Research Workshop, Nairobi, Kenya.
Chirwa, E.W. (2001), Liberalization of Food Marketing and Market Integration in Malawi,
Report of AERC Sponsored Research.
Cirera, X. & Arndt, C. (2006). ‘Measuring the impact of road rehabilitation on spatial market
efficiency in maize markets in Mozambique’. DNEAP discussion paper 30, Ministry of
Planning and Development, Mozambique.
Conforti, P. (2004). Price transmission in selected agricultural markets. FAO Commodity and
Trade Policy Research Working Paper No. 7. Commodities and Trade Division of the
Food and Agricultural Organization of the United Nations, Rome, Italy.
Cudjoe, G., Breisinger, C. & Diao, X. (2008). ‘Local impacts of a global crisis: Food price
transmission and poverty impacts in Ghana’. Discussion paper. International Food Policy
Research Institute.
Dercon, S. (1995). On market integration and liberalisation: method and application to
Ethiopia. The Journal of Development Studies, 32(1), 112-143.
University of Ghana http://ugspace.ug.edu.gh
88
Dickey, D. & Fuller, W. (1979).Distribution of the estimators for autoregressive time series with
a unit root.Journal of the American Statistical Association, 74, 427-431.
Dickey, D. & Fuller, W. (1981): Likelihood Ratio Statistics for Autoregressive Time Series with
a Unit Root. Econometrica, Vol. 49, pp. 1057 – 1072.
Enders, W. & Siklos, P. L. (2001).Cointegration and threshold adjustment.Journal of Business
and Economic Statistics 19, 166-176.
Enders, W., & Granger, C. W. J. (1998). Unit-root tests and asymmetric adjustment with an
example using the term structure of interest rates. Journal of Business & Economic
Statistics, 16(3), 304-311.
Engle, R. F. & Granger C. W. J. (1987). Cointegration and error correction: representation,
estimation and testing. Econometrica, 55, 251-276.
Engle, R. F., & Yoo, B. S. (1987). Forecasting and testing in co-integrated systems. Journal of
econometrics, 35(1), 143-159.
Enke, S. 1951. Equilibrium among spatially separated markets: solution by electrical analogue.
Econometrica. 19(7): 40.
FACKLER, P. (1996). Spatial price analysis: A methodological review. Mineo. North Carolina
State University.
FAO (2013). A value-chain analysis of international fish trade and food security with an impact
assessment of the small-scale sector. Summary Article, NORAD-FAO Project (January
2013). IIFET 2012, 10 p.
University of Ghana http://ugspace.ug.edu.gh
89
FASDEP I (2002): Food and Agriculture Sector Development Policy, Ministry of Agriculture,
Ghana. 55 FASDEP II (2008): Food and Agriculture Sector Development Policy,
Ministry of Agriculture, Ghana.
FAO (2016). Fishery and Aquaculture Country Profiles.Ghana (2016). Country Profile Fact
Sheets. In: FAO Fisheries and Aquaculture Department [online]. Rome. Updated
2016.[Cited 21 June 2018]. http://www.fao.org/fishery/
Fossati, S., Lorenzo, F., & Rodriguez, C. M. (2007). Regional and International Market
Integration of a Small Open Economy. Journal of Applied Economics, 10(1).
Garrow, J. S. & James, W. P. T. (1994).Human Nutrition and Dietetics. London, UK: Churchill
Livingstone.
Goletti, F. & Babu, S. 1994. Market liberalization and market integration of maize markets in
Malawi.Agriculture Economics. 11: 311-324.
Goletti, F., Ahmed, R. & Farid, N. (1995). Structural determinant of market integration: the case
study of rice in Bangladesh. The Development Economics, 32, 245-264.
Goodwin B. K. & Schroeder T. C. (1991). Co-integration test and spatial price linkages in
regional cattle markets. American Journal of Agricultural Economics, 73(2):452-464
Goodwin, B. K. & Piggott, N. E. (1999).‘Spatial market integration in the presence of threshold
effects’.American Agricultural Economics Association.
Gopal M. S., Shanmugam T.R., Raveendaran N., & Rajan G.K. (2009). Co integration and
Market Integration: An Application to Indian and International Black Tea Auction
Markets. Journal of Asian Business Management, 1(2), 193-205.
University of Ghana http://ugspace.ug.edu.gh
90
Gordon, A., Pulis, A., & Owusu-Adjei, E. (2011). Smoked marine fish from Western Region,
Ghana: a value chain assessment.
Gordon, D. V., & Maurice, S. (2015). Vertical and horizontal integration in the Uganda fish
supply chain: Measuring for feedback effects to fishermen. Aquaculture Economics &
Management, 19(1), 29-50.
Granger, C. W., & Lee, T. H. (1989). Investigation of production, sales and inventory
relationships using multicointegration and non‐ symmetric error correction
models. Journal of applied econometrics, 4(S1), S145-S159.
Gujarati, D. N. (2008). Basic Econometrics . McGraw-Hill Higher Education. International
Edition.
Gupta, S., & Mueller, R. A. (1982). Analyzing the pricing efficiency in spatial markets: Concept
and application. European Review of Agricultural Economics, 9(3), 301-312.
Hendry, D. F. (1986). Econometric modelling with cointegrated variables: an overview. Oxford
bulletin of economics and statistics, 48(3), 201-212.
Hossain, M. I. & Verbeke, W. (2010). Evaluation of rice markets integration in Bangladesh.The
Lahore Journal of Economics, 15 (2), 77-96.
Houck, J. P. (1977). An approach to specifying and estimating nonreversible functions.American
Journal of Agricultural Economics, 59, 570-572.
Isard, P. (1977). How far can we push the" law of one price"?. The American Economic
Review, 67(5), 942-948.
University of Ghana http://ugspace.ug.edu.gh
91
Johansen, S. & Juselius, K. (1990).Maximum likelihood estimation and inference on
cointegration with application to the demand for money. Oxford Bulletin of Economics
and Statistics, 52(2), 169-210.
Johansen, S. (1988). Statistical analysis of cointegration vectors. Journal of economic dynamics
and control, 12(2-3), 231-254.
Johansen, S. (1991). Estimation and hypothesis testing of cointegration vectors in Gaussian
vector autoregressive models. Econometrica: Journal of the Econometric Society, 1551-
1580.
Kanakaraj M., (2010), An empirical validity of market integration hypothesis with special
reference to the pepper market in Kerala. Pondicherry University, Dissertation. URL:
http://shodhganga.inflibnet.ac.in/handle/10603/900
Kawarazuka, N., & Béné, C. (2011). The potential role of small fish species in improving
micronutrient deficiencies in developing countries: building evidence. Public health
nutrition, 14(11), 1927-1938.
Kohls R.L. & Uhl J.N. (1998). Marketing of agricultural products. Fifth Edition, Macmillan, New
York.
Kovenock, D. & Widdows, K. (1998).Price leadership and asymmetric price rigidity. European
Journal of Political Economy, 14 (1), 167-187.
Kwiatkowski, D., Phillips, C. B., Schmidt, P., & Shin, Y. (1992). Testing the null hypothesis of
stationarity against the alternative of a unit root: how sure are we that economic time
series have a unit root? J. Econometrics 54, 159-178.
University of Ghana http://ugspace.ug.edu.gh
92
Langyintuo, A. S. (2010, September). Grain distribution in Ghana under imperfectly competitive
market conditions. In AAAE Third Conference/AEASA 48th Conference, Cape Town,
South Africa.
Lovendal, C. R., Jakobsen, K. T., & Jacque, A. (2007).Food Prices and Food Security in Trinidad
and Tobago.Agricultural Development Economics Division.Food and Agricultural
Organization. ESA Working paper No. 0727..
Mafimisebi, T. E. (2001). Spatial price equilibrium and fish market integration in
Nigeria. Unpublished PhD thesis. University of Ibadan, Nigeria, 201.
Mafimisebi, T. E. (2012). Spatial equilibrium, market integration and price exogeneity in dry fish
marketing in Nigeria: a vector auto-regressive approach. Journal of Economics, Finance
and Administrative Science, 17 (33), 31-37,
McCorriston, S., Morgan, C. W., & Rayner, A. J. (1998). Processing technology, market power
and price transmission. Journal of Agricultural Economics, 49(2), 185-201.
Mensah, C. A. (2012). Optimisation of profit in the artisanal marine fishing: a case study of
Sekondi fishing harbour(Doctoral dissertation).
Mensah-Bonsu, A., Agyeiwaa-Afran, A. & Kuwornu, J. K. M. (2011). Efficiency of the plantain
marketing system in Ghana: a co-integration analysis. Journal of Development and
Agricultural Economics, 3(12), 593–601.
Meyer, J. & Von Cramon-Taubadel, S. (2004): Asymmetric price transmission: a survey, Journal
of Agricultural Economics, 55(3), 581-611.
MOFA (2016). ‘Agricultural Sector Progress Report 2016’. Accra, Ghana
University of Ghana http://ugspace.ug.edu.gh
93
Motamed, M., Foster, K. A., & Tyner, W. E. (2008). Applying cointegration and error correction
to measure trade linkages: maize prices in the United States and Mexico. Agricultural
Economics, 39(1), 29-39.
Negassa, A., & Myers, R. J. (2007). Estimating policy effects on spatial market efficiency: An
extension to the parity bounds model. American Journal of Agricultural
Economics, 89(2), 338-352.
Negassa, A., Meyers, R. & Gabre-Maldhin, E. (2003).Analyzing the grain market efficiency in
developing countries: review of existing methods and extensions to the parity bound
model. Discussion Paper, 63. Market Trade and Institutions Division.
Nielsen, M., Setala, J., Laitinen, J., Saarni, K., Virtanen, J., & Honkanen, A. (2006). Market
integration and causality in demand: the case study of farmed rout in Germany. In Paper
delivered at the 13th Biennial Conference of International Institute of Fisheries
Economics and Trade, Portsmouth, UK, July (pp. 11-14).
Nkang, M. N., Hentry, M.N. & Gabriel, N.O. (2007) Price transmission and integration of cocoa
and palm oil markets in cross river state, Nigeria: Implications for rural development,
Agricultural Journal, 2(4): 457-463.
Norman-Lopez, A., & Asche, F. (2008).Competition between imported tilapia and US catfish in
the US market. Marine Resource Economics, 23(2), 199-214.
Obodai, E.A., Muhammad, B.A., Obodai, G.A., & Opoku, E. (2009). “Effect of fuel wood on the
quality of smoked freshwater fish species sold in the Tamale Central Market, Northern
University of Ghana http://ugspace.ug.edu.gh
94
Region. Ghana.” Ethiopian journal of environmental studies and management volume
2.No 2.2009
Ofori-Danson, P. K., Sarpong, D. B., Sumaila, U. R., Nunoo, F. K., & Asiedu, B. (2013). Poverty
measurements in small-scale fisheries of Ghana: A step towards poverty
eradication. Journal: Current Research Journal of Social Sciences, 5, 75-90.
Ofori-Danson, P. K., Sarpong, D. B., Sumaila, U. R., Nunoo, F. K., & Asiedu, B. (2013). Poverty
measurements in small-scale fisheries of Ghana: A step towards poverty
eradication. Journal: Current Research Journal of Social Sciences, 5, 75-90.
Okon, R. N., & Egbon, P. C. (2003). The Integration of Nigeria’s Rural and Urban Food Stuff
Markets. A Report of African Economic Research Consortium (AERC) Sponsored
Research, 1-17.
Oladapo, O. O. (2003). Market integration for pineapples in Nigeria. An Unpublished Ph. D
Thesis, University of Agriculture, Abeokuta, Nigeria.
Onyuma, S. O., Icart, E., & Owuor, G. (2006, August). Testing market integration for fresh
pineapples in Kenya. In International Association of Agricultural Economics Conference,
Gold Coast, Australia, August (pp. 12-18).
Peltzman, S. (2000). Prices rise faster than they fall. Journal of political economy, 108(3), 466-
502.
Phillips, P., & P. Perron. 1988. “Testing for a Unit Root in Time Series Regression.” Biometrika
75 (2): 335–346.
University of Ghana http://ugspace.ug.edu.gh
95
Prajogo, D., & Olhager, J. (2012). Supply chain integration and performance: The effects of long-
term relationships, information technology and sharing, and logistics
integration. International Journal of Production Economics, 135(1), 514-522.
Quaas, M. F., Hoffmann, J., Kamin, K., Kleemann, L., & Schacht, K. (2016). Fishing for
Proteins. How marine fisheries impact on global food security up to 2050. A global
prognosis.
Rapsomanikis, G. & Karfakis, P. (2007). ‘Margins across time and space: threshold cointegration
and spatial pricing applications to commodity markets in Tanzania’. Paper presented in
the workshop on staple food trade and market policy options for promoting development
in eastern and southern Africa, Rome.
Rapsomanikis, G., Hallam, D., & Conforti, P. (2006). Market integration and price transmission
in selected food and cash crop markets of developing countries: review and
applications. Agricultural Commodity Markets and Trade, 187-217.
Rashid S., 2004. Spatial Integration of Maize Markets in Post-Liberalised Uganda. Markets,
Trade and Institutions Division (MTID) of IFPRI Discussion Paper No. 71.
Ravallion, M. (1986).Testing market integration.American Journals for Agriculture Economics,
1, 102-108.
Reagan, P. B., & Weitzman, M. L. (1982). Asymmetries in price and quantity adjustments by the
competitive firm. Journal of Economic Theory, 27(2), 410-420.
Robadue, D., Torell, E., McNally, C., Kaine, K., & Lazar, N. (2018). Summary of the 2016-2018
directed study course for Graduate Students studying integrated fisheries
University of Ghana http://ugspace.ug.edu.gh
96
management. The USAID/Ghana Sustainable Fisheries Management Project (SFMP).
Narragansett, RI: Coastal Resources Center, Graduate School of Oceanography,
University of Rhode Island. GH2014_SCI045_CRC.
Rico, I. (2009). ‘Models for analyzing nonlinearities in price transmission’. (A PhD dissertation,
Faculty of Agricultural Sciences, Georg-August-university, Göttingen, Germany, 2009).
Samey, B. (2015). Literature Review On Sardinella Value Chain. The USAID/Ghana Sustainable
Fisheries Management Project (SFMP). Narragansett, RI: Coastal Resources Center,
Graduate School of Oceanography, University of Rhode Island and SNV Netherlands
Development Organization. GH2014_SCI007_SNV.
Samuelson, P. 1952. Spatial price equilibrium and linear programming.American Economic
Review. 42: 283–303.
Sapkota, P., Dey, M. M., Alam, M. F., & Singh, K. (2012). Price transmission relationships along
the seafood value chains in Bangladesh: aquaculture and capture fishery species.
Seneshaw T (2013). Spatial integration of cereal markets in Ethiopia, Ethiopia Strategy Support
Program, Ethiopian Development Research Institute, ESSP Working Paper 56.
Sexton, R. J., King, C. L. & Carman, H. F. (1991). Market integration, efficiency of arbitrage,
and imperfect competition: methodology and application to U.S celery. American Journal
of Agricultural Economics, 73(3), 568-80.
Shinoj, P., Kumar, B. G., Sathiadhas, R., Datta, K. K., Menon, M., & Singh, S. K. (2008). Spatial
price integration and price transmission among major fish markets in India. Agricultural
Economics Research Review, 21(2008).
University of Ghana http://ugspace.ug.edu.gh
97
Shively, G. E. (1996). Food price variability and economic reform: an arch approach for Ghana.
American Journal of Agricultural Economics, 78 (1), 126–136.
Spiller, P. T., & Huang, C. J. (1986). On the extent of the market: wholesale gasoline in the
northeastern United States. The Journal of Industrial Economics, 131-145.
Spiller, P. T., & Wood, R. O. (1988). The estimation of transaction costs in arbitrage
models. Journal of Econometrics, 39(3), 309-326.
Stigler, G.J. (1969). The Theory of Price. 85 pp. London: Macmil
Takayama, T. & Judge, G. 1971.Spatial and Temporal Price Allocation Models.North-Holland.
Amsterdam.
Tomek, W. G. & Robinson, K. L. (2003).Agricultural product prices. Ithaca, NY: Cornell
University Press.
Tong, H. (1978). On a threshold model in pattern recognition and signal processing,(Ed) C. Chen.
Sijhoff and Noonhoff, Amsterdam.
Van Campenhout, B. (2007). Modelling trends in food market integration: Method and an
application to Tanzanian maize markets. Food Policy, 32(1), 112-127.
Von Cramon-Taubadel, S. (1998). Estimating asymmetric price transmission with the error
correction representation: an application to the German pork market. European Review of
Agricultural Economics, 25, 1-18.
University of Ghana http://ugspace.ug.edu.gh
98
Von Cramon‐ Taubadel, S., & Loy, J. P. (1996). Price asymmetry in the international wheat
market: Comment. Canadian Journal of Agricultural Economics/Revue canadienne
d'agroeconomie, 44(3), 311-317.
Ward, R. W. (1982). Asymmetry in retail, wholesale and shipping point pricing for fresh
vegetables. American Journal of Agricultural Economics 62, 205-212.
Wlazlowski, S., Giulietti, M., Binner, J. & Milas, C. (2009).Price dynamics in European
petroleum markets. Energy Economics, 31(1), 99-108.
Wolffram, R. (1971). Positivistic measures of aggregate supply elasticities: Some new
approaches: Some critical notes. American Journal of Agricultural Economics, 53(2),
356-359.
Wyeth, J. (1992). The measurement of market integration and applications to food security
policies. Sussex Univ
University of Ghana http://ugspace.ug.edu.gh
99
APPENDICES
APPENDIX 1: NOMINAL KOOBI PRICE DATA (2012-2017) FROM GSS
Year Cape Coast Ada Agbozume Half Assini Kpandu Mankessim Techiman Tema Wa Makola
2012m1 8.5 10.5 10.8 10.5 10.8 11.8 11.0 11.0 11.1 11.8
2012m2 8.5 10.5 10.8 10.5 10.8 12.5 12.5 11.0 12.5 12.5
2012m3 8.5 10.5 10.8 10.5 10.8 12.5 12.5 11.0 11.1 12.5
2012m4 8.5 10.5 10.8 10.5 10.8 12.7 12.5 12.4 11.1 12.7
2012m5 8.5 10.7 10.8 10.5 10.8 12.7 12.5 12.4 11.1 12.7
2012m6 8.5 10.7 10.8 10.5 10.8 12.9 12.5 12.4 11.1 12.9
2012m7 8.5 10.7 10.8 10.5 10.8 13.1 12.8 12.4 11.1 13.1
2012m8 8.5 10.7 10.8 10.5 10.8 12.7 12.8 12.4 11.6 12.7
2012m9 9.5 10.7 10.8 10.5 12.8 11.8 12.8 12.4 11.6 11.8
2012m10 9.5 11.3 10.8 10.9 12.8 12.4 12.8 12.4 11.6 12.4
2012m11 9.5 11.3 10.8 10.9 12.8 12.5 12.8 12.4 11.6 12.5
2012m12 9.5 11.3 11.1 10.9 12.8 12.7 12.8 12.4 11.6 12.7
2013m1 9.8 11.3 11.7 10.9 12.8 12.7 12.8 12.7 12.0 13.0
2013m2 9.8 11.5 11.7 10.9 12.8 12.7 12.8 12.7 12.0 13.0
2013m3 9.8 11.5 11.8 10.9 12.8 12.7 12.8 13.0 12.0 13.0
2013m4 9.8 11.8 11.8 10.9 12.8 12.9 13.0 13.0 12.0 13.4
2013m5 9.8 11.8 11.8 10.9 12.8 12.9 13.0 13.0 12.0 13.4
2013m6 9.8 11.8 11.8 11.7 12.8 12.9 13.0 13.0 12.0 13.4
2013m7 9.8 11.8 11.8 11.7 12.8 12.9 13.0 13.0 12.0 13.6
2013m8 9.8 12.3 11.8 11.7 12.8 12.9 13.0 13.0 12.0 13.3
2013m9 9.8 12.3 11.8 11.7 12.8 12.9 13.0 13.0 12.0 12.8
2013m10 9.8 12.3 11.8 11.7 12.8 12.9 13.0 13.0 12.3 12.7
2013m11 9.8 12.3 11.8 11.7 12.8 12.9 13.0 13.0 12.3 12.9
2013m12 9.8 12.3 12.0 11.7 12.8 12.9 13.0 13.0 12.3 12.4
2014m1 9.8 12.4 12.0 11.7 12.8 13.2 13.0 13.0 12.3 14.0
2014m2 11.8 12.4 12.0 11.7 12.8 13.2 13.0 13.0 12.3 14.1
University of Ghana http://ugspace.ug.edu.gh
100
2014m3 11.8 12.4 12.6 11.7 12.8 13.2 13.0 13.0 12.3 14.4
2014m4 11.8 12.9 12.6 12.3 12.8 13.2 13.0 13.3 12.3 13.8
2014m5 11.8 12.9 12.9 12.3 12.8 13.2 13.3 13.3 13.0 14.0
2014m6 11.9 12.9 12.9 12.3 12.8 13.2 13.3 13.3 13.0 14.1
2014m7 11.9 12.9 12.9 12.3 12.8 13.2 13.3 13.3 13.0 14.6
2014m8 11.9 12.9 12.9 12.3 12.8 14.2 13.3 13.3 13.0 14.2
2014m9 11.9 12.9 12.9 12.3 12.8 14.2 13.3 13.7 13.0 14.2
2014m10 11.9 13.3 12.9 12.3 22.2 14.2 13.3 13.7 13.0 14.2
2014m11 12.2 13.3 12.6 13.8 12.8 14.3 13.3 13.7 13.0 14.3
2014m12 12.2 13.3 12.6 13.8 12.8 14.3 13.3 13.7 13.0 13.9
2015m1 12.2 13.3 13.3 13.8 12.8 14.3 13.8 13.7 13.0 15.1
2015m2 12.2 13.8 13.3 13.8 12.8 14.7 13.8 14.2 13.0 15.2
2015m3 12.2 13.8 13.3 13.8 12.8 15.3 13.8 14.2 13.7 15.3
2015m4 12.2 13.8 13.3 14.7 12.8 15.3 13.8 14.2 13.7 15.8
2015m5 12.2 13.9 14.3 14.7 12.8 15.3 13.8 14.2 13.7 15.9
2015m6 13.3 14.2 14.3 14.7 12.8 15.3 15.0 14.2 13.7 16.2
2015m7 13.3 14.2 14.3 14.7 12.8 15.3 15.0 14.2 13.7 16.0
2015m8 13.3 14.2 14.9 14.7 12.8 15.3 15.0 14.2 14.2 15.9
2015m9 13.3 14.2 14.9 15.2 12.8 15.3 15.0 14.2 14.2 15.1
2015m10 13.3 14.2 14.9 15.2 12.8 15.3 15.0 14.8 14.2 15.1
2015m11 13.3 14.6 14.9 15.2 12.8 15.3 15.0 14.8 14.2 15.3
2015m12 13.3 14.6 14.9 15.2 12.8 15.3 15.0 14.8 14.2 15.3
2016m1 13.3 14.6 14.9 15.2 12.8 16.6 15.0 14.8 14.2 16.6
2016m2 13.3 14.6 14.9 15.2 12.8 16.6 15.3 14.8 14.2 16.8
2016m3 13.3 15.5 14.9 15.2 12.8 16.6 15.3 14.8 14.2 16.8
2016m4 13.5 15.5 15.6 15.8 12.8 16.6 15.3 14.8 14.2 17.1
2016m5 13.5 16.2 15.6 15.8 12.8 16.6 15.3 14.8 14.2 17.5
2016m6 13.5 16.2 15.6 15.8 12.8 16.6 15.3 14.8 14.2 17.7
2016m7 13.5 16.2 15.6 15.8 12.8 16.6 15.3 14.8 14.2 17.5
2016m8 13.5 16.6 15.6 15.8 12.8 16.6 15.3 14.8 15.2 17.1
2016m9 13.5 16.6 15.6 15.8 12.8 16.6 15.3 14.8 15.2 16.6
University of Ghana http://ugspace.ug.edu.gh
101
2016m10 13.5 16.6 15.6 15.8 12.8 16.6 15.9 14.8 15.2 16.5
2016m11 13.5 16.6 16.2 16.2 12.8 16.6 15.9 14.8 15.2 17.1
2016m12 13.5 16.6 16.2 16.2 12.8 16.6 15.9 14.8 15.2 17.8
2017m1 14.3 16.6 16.2 16.2 12.8 18.8 17.1 14.8 15.2 18.8
2017m2 14.3 16.6 16.2 16.2 12.9 18.8 17.1 14.8 15.2 19.0
2017m3 14.3 16.6 16.2 16.2 12.9 18.8 17.1 14.8 15.2 19.2
2017m4 14.3 16.6 17.5 16.2 12.9 18.8 17.1 14.8 15.2 19.4
2017m5 14.3 16.6 17.5 16.2 12.9 18.8 17.1 14.8 15.2 19.3
2017m6 14.3 16.6 17.5 16.2 12.9 18.8 17.1 14.8 17.0 19.6
2017m7 14.3 16.6 17.5 16.2 12.9 18.8 17.1 14.8 17.0 19.6
2017m8 14.3 17.6 17.5 16.2 12.9 18.8 17.8 15.2 17.0 19.6
2017m9 14.3 17.6 17.5 16.2 12.9 18.8 17.8 15.2 17.0 19.2
2017m10 15.0 17.6 17.5 16.2 12.9 18.8 17.8 15.2 17.0 19.2
2017m11 15.4 17.6 17.5 18.0 12.9 18.8 17.8 15.2 17.0 19.4
2017m12 15.4 17.6 17.8 18.0 13.3 18.8 17.8 15.2 17.0 19.6
University of Ghana http://ugspace.ug.edu.gh
102
APPENDIX 2: NOMINAL KPALA PRICE DATA (2012-2017) FROM GSS
Year Cape Coast Ada Half Assini Mankessim Techiman Tema Wa Makola
2012m1 16.5 8.9 10.1 10.6 10.1 9.6 8.3 10.6
2012m2 16.5 9.1 10.1 11.0 10.1 9.6 8.3 10.6
2012m3 16.5 9.3 10.1 11.0 10.4 10.6 9.3 10.6
2012m4 16.5 9.3 10.1 11.5 10.4 10.6 9.3 11.2
2012m5 16.5 9.3 10.1 11.5 10.4 10.6 9.3 11.2
2012m6 16.5 9.7 10.1 11.5 10.4 10.6 9.9 11.5
2012m7 16.5 10.2 10.1 11.5 10.4 10.6 9.9 11.5
2012m8 16.5 10.2 10.1 11.5 10.4 10.6 10.4 11.5
2012m9 16.5 10.2 10.1 11.9 10.8 11.0 10.4 11.9
2012m10 16.5 10.2 10.1 11.8 10.8 11.0 10.4 11.8
2012m11 16.5 10.5 10.1 11.5 10.8 11.0 10.4 11.5
2012m12 16.5 10.5 10.1 11.7 10.8 11.0 10.4 11.7
2013m1 16.5 10.5 10.5 12.0 10.8 11.4 10.8 12.0
2013m2 16.5 10.7 10.5 12.0 10.8 11.4 10.8 12.0
2013m3 16.5 10.7 10.5 12.0 10.8 11.4 10.8 12.0
2013m4 16.5 11.7 10.5 12.0 10.8 11.4 10.8 12.3
2013m5 16.5 11.7 10.5 12.3 10.8 11.4 10.8 12.3
2013m6 17.1 11.7 11.2 12.3 10.8 11.4 10.8 12.3
2013m7 17.1 12.2 11.2 12.4 11.0 11.4 11.4 12.3
2013m8 17.1 12.2 11.2 12.4 11.0 11.4 11.4 12.3
2013m9 17.1 12.2 11.2 12.4 11.0 11.4 11.4 12.4
2013m10 17.1 12.2 11.6 12.4 11.0 11.9 11.4 12.4
2013m11 17.1 12.2 11.6 12.4 11.0 11.9 11.4 12.4
University of Ghana http://ugspace.ug.edu.gh
103
2013m12 17.1 12.2 11.6 12.4 11.0 11.9 11.4 12.4
2014m1 17.1 12.4 11.6 12.5 11.3 11.9 11.4 12.5
2014m2 17.1 12.4 11.9 12.5 11.3 12.1 11.4 12.5
2014m3 17.1 12.4 11.9 12.5 11.3 12.1 11.4 12.5
2014m4 17.1 12.4 11.9 12.5 11.3 12.1 11.4 12.5
2014m5 17.1 12.4 11.9 12.5 11.5 12.1 11.9 12.9
2014m6 17.1 12.4 11.9 12.7 11.5 12.1 11.9 12.9
2014m7 17.1 12.4 11.9 12.7 11.5 12.4 11.9 12.9
2014m8 18.3 12.6 11.9 12.9 11.5 12.4 11.9 12.9
2014m9 18.3 12.6 11.9 12.9 11.5 12.4 11.9 12.9
2014m10 18.3 12.6 11.9 12.9 11.5 12.4 11.9 12.9
2014m11 18.7 12.6 11.9 12.9 11.9 12.4 11.9 12.9
2014m12 18.7 12.6 11.9 12.9 11.9 12.4 11.9 12.9
2015m1 18.7 12.6 11.9 12.9 11.9 12.4 12.3 12.5
2015m2 18.7 12.6 11.9 12.9 11.9 12.4 12.3 12.5
2015m3 18.7 12.6 11.9 12.9 11.9 12.4 12.3 12.5
2015m4 18.7 12.6 11.9 12.9 11.9 12.4 12.3 12.5
2015m5 18.7 12.6 11.9 12.9 11.9 12.4 12.3 12.5
2015m6 18.7 12.6 11.9 12.9 11.9 12.4 12.3 12.5
2015m7 18.7 12.6 11.9 12.9 11.9 12.4 12.3 12.5
2015m8 18.7 12.6 11.9 12.9 11.7 12.4 12.3 12.5
2015m9 18.7 12.6 11.9 12.9 11.4 12.4 12.3 12.5
2015m10 18.7 12.6 11.9 12.9 11.7 12.4 12.3 12.5
2015m11 18.7 12.6 11.9 12.9 11.7 12.4 12.3 12.5
2015m12 18.7 12.6 11.9 12.9 11.5 12.4 12.3 12.5
2016m1 18.7 12.7 12.0 12.9 11.8 12.6 12.3 12.7
University of Ghana http://ugspace.ug.edu.gh
104
2016m2 18.7 12.7 12.0 12.9 12.1 12.6 12.3 12.7
2016m3 18.7 12.7 12.0 12.9 12.1 12.6 12.3 12.7
2016m4 18.7 12.7 12.3 12.9 12.1 12.6 12.4 12.7
2016m5 18.7 12.7 12.3 12.9 12.1 12.6 12.4 12.9
2016m6 18.7 12.7 12.3 12.9 12.1 12.6 12.4 12.9
2016m7 18.7 12.7 12.3 12.9 12.1 12.6 12.4 12.9
2016m8 18.7 12.9 12.3 12.9 12.3 12.6 12.4 12.9
2016m9 18.7 12.9 12.3 12.9 12.3 12.8 12.4 12.9
2016m10 18.7 12.9 12.3 12.9 12.3 12.8 12.4 12.9
2016m11 18.7 12.9 12.5 13.0 12.3 12.8 12.4 13.0
2016m12 18.7 12.9 12.5 13.0 12.3 12.8 12.4 13.0
2017m1 18.9 12.9 13.5 13.0 12.5 12.8 12.4 13.0
2017m2 18.9 12.9 13.5 13.0 12.5 12.8 12.4 13.0
2017m3 18.9 13.0 13.4 13.4 12.5 12.8 12.6 13.4
2017m4 18.9 13.0 13.4 13.4 12.5 12.8 12.6 13.4
2017m5 18.9 13.0 13.4 13.4 12.5 12.8 12.6 13.4
2017m6 18.9 13.0 13.4 13.4 12.8 12.8 12.6 13.4
2017m7 18.9 13.0 13.4 13.4 12.8 13.2 12.6 13.4
2017m8 18.9 13.0 13.4 13.4 12.8 13.2 12.6 13.4
2017m9 18.9 13.0 13.4 13.4 12.8 13.2 12.9 13.4
2017m10 18.9 13.0 13.4 13.4 13.1 13.2 12.9 13.4
2017m11 18.9 13.0 13.6 13.7 13.1 13.2 12.9 13.7
2017m12 18.9 13.0 13.6 13.7 13.1 13.2 12.9 13.7
University of Ghana http://ugspace.ug.edu.gh
105
APPENDIX 3: NOMINAL SMOKED HERRING PRICE DATA (2012-2017) FROM GSS
Year CapeCoast Ada Agbozume HalfAssini Kpandu Mankessim Techiman Tema Wa Makola
2012m1 9.7 9.6 10.5 10.8 9.6 11.9 10.3 10.3 10.6 11.9
2012m2 9.7 9.6 10.5 10.8 9.6 12.2 10.6 10.3 10.6 12.2
2012m3 9.7 9.6 10.5 10.8 9.6 12.2 11.3 10.3 10.6 12.2
2012m4 9.7 10.6 10.5 10.8 9.6 12.6 11.3 10.3 10.6 12.6
2012m5 9.7 10.9 10.5 10.8 9.6 12.6 11.3 10.3 10.6 12.6
2012m6 9.7 11.4 10.5 10.8 9.6 12.7 11.3 11.1 11.3 12.7
2012m7 9.7 11.4 10.5 10.8 9.6 12.8 11.3 11.1 11.3 12.8
2012m8 9.7 11.4 10.5 10.8 9.6 12.5 11.3 11.1 11.3 12.5
2012m9 9.7 11.4 11.2 10.8 10.8 11.9 11.3 11.1 11.3 11.9
2012m10 9.7 11.4 11.2 10.8 16.2 12.0 11.3 11.1 11.3 12.0
2012m11 9.7 11.4 11.2 10.8 10.9 12.2 11.3 11.1 11.3 12.2
2012m12 9.7 11.4 11.6 10.8 10.9 12.2 11.7 11.1 11.3 12.2
2013m1 9.7 11.7 11.6 11.3 10.9 12.8 11.7 12.1 11.7 13.0
2013m2 9.7 11.7 11.6 11.3 10.9 12.8 11.7 12.1 11.7 13.0
2013m3 9.7 11.7 12.3 11.3 10.9 12.8 11.7 12.1 11.7 13.0
2013m4 9.7 12.1 12.3 11.3 10.9 12.8 11.7 12.1 11.7 13.2
2013m5 9.7 12.1 12.4 11.3 11.4 12.8 11.7 12.1 12.2 13.2
2013m6 9.7 12.1 12.4 11.3 11.4 12.8 11.7 12.1 12.2 13.2
2013m7 10.2 12.1 12.4 11.3 11.4 12.8 12.4 12.1 12.2 13.6
2013m8 10.2 12.6 12.4 11.3 11.4 12.8 12.4 12.1 12.2 13.4
2013m9 10.2 12.6 12.4 11.3 11.4 12.8 12.4 12.1 12.2 12.9
2013m10 10.2 12.6 12.4 11.5 12.1 12.8 12.4 12.1 12.2 12.8
2013m11 10.2 12.6 12.4 11.5 12.1 12.8 12.4 12.1 12.2 12.9
2013m12 10.2 12.6 12.4 11.5 12.1 12.8 12.4 12.1 12.2 12.8
2014m1 10.2 12.9 12.4 11.5 12.1 13.3 12.4 12.1 12.2 13.7
2014m2 10.2 12.9 12.4 11.8 12.1 13.3 12.4 12.9 12.2 13.8
2014m3 10.2 12.9 12.4 11.8 12.5 13.3 12.4 12.9 12.2 13.8
2014m4 10.2 13.3 13.5 11.8 12.5 13.3 12.4 12.9 12.2 14.1
University of Ghana http://ugspace.ug.edu.gh
106
2014m5 10.8 13.3 13.5 11.8 12.5 13.3 12.9 12.9 12.9 14.1
2014m6 10.9 13.7 13.5 12.7 12.5 13.3 12.9 12.9 12.9 14.2
2014m7 10.9 13.7 13.5 12.7 12.5 13.3 12.9 12.9 12.9 14.5
2014m8 11.3 13.7 14.1 12.7 12.5 14.4 12.9 12.9 12.9 14.4
2014m9 11.3 13.7 14.1 12.7 12.5 14.4 12.9 12.9 12.9 14.4
2014m10 11.3 13.9 14.1 12.7 12.5 14.4 12.9 12.9 13.6 14.4
2014m11 11.4 13.9 14.1 12.7 13.5 14.5 13.6 12.9 13.6 14.5
2014m12 11.4 13.9 14.1 12.7 7.6 14.5 13.6 12.9 13.6 14.1
2015m1 12.0 13.9 14.1 12.7 13.8 14.5 13.6 12.9 13.9 15.2
2015m2 12.1 14.2 14.1 12.7 13.8 14.5 13.6 13.3 13.9 15.3
2015m3 12.2 14.2 14.1 12.7 14.0 15.4 13.6 13.3 13.9 15.4
2015m4 12.2 14.2 14.1 12.9 14.0 15.4 13.6 13.3 13.9 15.5
2015m5 12.2 14.5 14.1 12.9 14.0 15.4 13.6 13.3 13.9 15.7
2015m6 12.4 14.5 14.1 12.9 14.7 15.4 13.9 13.3 13.9 16.0
2015m7 12.2 14.5 14.1 13.3 14.7 15.4 13.9 13.3 13.9 15.9
2015m8 12.3 14.5 14.8 13.3 14.7 15.4 13.9 13.3 14.5 15.6
2015m9 12.0 14.5 14.8 13.3 14.7 15.4 13.9 13.3 14.5 15.3
2015m10 12.0 14.5 14.8 13.3 14.7 15.4 13.9 13.8 14.5 15.3
2015m11 12.1 15.1 14.8 13.3 14.7 15.4 13.9 13.8 14.5 15.4
2015m12 12.2 15.1 16.1 14.3 15.4 15.4 13.9 13.8 14.5 15.4
2016m1 12.7 15.1 14.8 14.3 15.4 16.4 15.2 13.8 14.5 16.4
2016m2 12.7 15.1 14.8 14.3 15.4 16.4 15.2 13.8 14.5 16.4
2016m3 12.7 15.1 14.8 14.3 15.4 16.4 15.2 13.8 14.5 16.4
2016m4 12.9 15.7 15.3 14.3 15.4 16.4 15.2 13.8 14.5 16.6
2016m5 13.1 15.7 15.3 14.3 15.4 16.4 15.2 14.3 14.5 16.9
2016m6 13.3 15.7 15.3 14.3 15.4 16.4 15.2 14.3 14.5 17.2
2016m7 13.1 15.9 15.3 14.3 15.4 16.4 15.2 14.3 14.5 16.8
2016m8 13.3 16.2 15.3 14.8 15.4 16.4 15.2 14.7 14.5 16.6
2016m9 13.1 16.1 15.3 14.8 15.4 16.4 15.2 14.7 14.5 15.5
2016m10 13.1 16.2 15.3 14.8 15.4 16.4 15.2 14.7 14.5 15.5
University of Ghana http://ugspace.ug.edu.gh
107
2016m11 13.5 16.2 16.1 15.6 15.4 16.4 15.2 14.7 14.5 17.4
2016m12 13.5 16.2 16.1 15.6 15.4 16.4 15.2 14.7 14.5 17.4
2017m1 14.6 16.8 16.1 15.6 15.4 18.5 15.2 15.7 14.5 18.5
2017m2 14.6 16.8 16.1 15.6 15.4 18.5 15.2 15.7 14.5 18.7
2017m3 14.6 16.8 16.1 17.6 15.4 18.5 15.2 15.7 16.2 18.7
2017m4 14.6 16.8 17.1 17.6 15.7 18.5 15.2 15.7 16.2 18.7
2017m5 14.6 16.8 17.1 17.6 15.7 18.5 15.2 15.7 16.2 18.7
2017m6 15.1 17.3 17.1 17.6 15.7 18.5 15.2 15.7 16.2 19.0
2017m7 15.1 17.3 18.7 17.6 15.7 18.5 15.2 15.7 16.2 19.0
2017m8 15.1 17.3 18.7 17.6 15.7 18.5 15.2 15.7 16.2 19.0
2017m9 15.1 17.3 18.7 17.6 15.7 18.5 15.2 15.7 16.2 18.6
2017m10 15.1 17.3 18.7 17.6 15.7 18.7 15.2 15.7 16.2 18.7
2017m11 15.1 17.3 18.7 17.6 15.7 18.7 15.2 15.7 17.4 19.0
2017m12 15.3 17.3 18.7 17.6 15.7 18.7 15.2 15.7 17.4 19.2
University of Ghana http://ugspace.ug.edu.gh
110
University of Ghana http://ugspace.ug.edu.gh
111
University of Ghana http://ugspace.ug.edu.gh